D1 by huanghengdong



       Deployment of


        Deliverable 1:




       15th July 1996
150796                                                               DIATS/TRG/WP100

CONTENTS                                                                       PAGE

EXECUTIVE SUMMARY                                                                  1

1.       INTRODUCTION                                                              2

         1.1   Motorway Congestion                                                 2
         1.2   Approach                                                            2
         1.3   Technology Based Opportunities and Functionality                    3

2.       STATE OF THE ART IN THE EU                                                5

         2.1   Cooperative driving systems                                         5
         2.2   Vehicle based autonomous systems                                    7
               2.2.1 Collision Avoidance Systems (CAS)                             7
               2.2.2 Vision Enhancement                                           12
               2.2.3 Driver Status Monitoring and Support                         14
                       (incorporating Automatic Enforcement)
               2.2.4 Dialogue Management                                          16
         2.3   Infrastructure based                                               17
               2.3.1 On-line speed control Systems                                17
               2.3.2 AID                                                          20
               2.3.3 Other                                                        22
               2.3.4 High Occupancy Vehicle Lanes (HOV)                           24
               2.3.5 Ramp Metering                                                26

3.       REVIEW OF DEVELOPMENTS IN THE U.S.A. AND JAPAN                           29

         3.1   U.S.A.                                                             29
         3.2   Japan                                                              31

4.       USERS AND PROVIDERS                                                      34

         4.1   Identification of User and Provider Groups                         34
         4.2   Issues and Future Trends                                           35

5.       CONCLUDING COMMENTS                                                      38

6.       REFERENCES                                                               39

ANNEX A:       Extracts from CORD Recommendations of Transport Telematics
               Functions and Subfunctions.                                        47

ANNEX B:       Glossary of Selected Abbreviations                                 51

DIATS                                    Deliverable 1
150796                                                 DIATS/TRG/WP100

Information on Deliverable 1:

Workpackage Leader and Deliverable Coordinator:

TRG Southampton

Primary Reviewer:


Contributions from:

TRG Southampton (Prime Contractor)
Heusch Boesefeldt

DIATS                                  Deliverable 1
150796                                                                        DIATS/TRG/WP100


This deliverable identifies a list of potential ATT systems which relate to base field tests of
'cooperative driving' measures on inter and peri urban roads. It is based on project tasks 110-130
which were completed in months 1-3 of this project, and provides a background for use in
producing benchmark scenarios for subsequent studies.

The scope of the systems examined has been based on those identified in the description of task
7.3/20 on which this project is based (DGVII, 1994), namely:

         ".... from improved information for drivers, warnings of hazards posed by other
         vehicles ... to "automatic driving" facilities .... that will allow automatic driving
         devices to take over some of the driver's responsibilities".

This deliverable has built on the existing knowledge of the members of the project consortium, and
activities performed as part of earlier CEC projects and commercial ventures (eg. CORD, 1996,
Fisher et. al., 1994, Transport Tech. 1996). Ten different types of systems have been identified in
the review process which may have a direct bearing on the evolution and deployment of ATT
systems ie:-
*        Cooperative driving systems,
*        Vehicle based autonomous systems including:-
         -        Collision Avoidance Systems (CAS),
         -        Vision Enhancement,
         -        Driver Status Monitoring and Support (incorporating Automatic Enforcement),
         -        Dialogue Management,
*        Infrastructure based:-
         -        On-line speed control Systems,
         -        AID,
         -        Other Infrastructure based,
         -        High Occupancy Vehicle Lanes (HOV),
         -        Ramp Metering.

The discussion of each system provides an overview of operational features, objectives and current
developments, as well as identifying relevant issues which may effect deployment. The review also
considers parallel developments being undertaken in the U.S.A. (eg. Harding et. al. 1995) and
Japan where a number of 'evolutionary paths' have already been identified. An element of foresight
has been applied to all the systems studied. This includes the potential for fundamental changes in
aspects such as driver behaviour.

In conclusion, this deliverable has established and defined the current state of the art of ATT
systems as they apply to the project, which will be further refined in later workpackage activities.

DIATS                                      Deliverable 1

150796                                                                         DIATS/TRG/WP100


This section sets the framework for the review of the candidate systems described in this


Congestion on urban and interurban roads has grown substantially in the last decade because of
increased private vehicle ownership, dependence and use. Despite a recent slowing of the growth
rate due to economic recession, forecasts still show a rise in car ownership and in vehicle kilometres
over the next 10-20 years (Bangemann et. al, 1994). New road construction will be unable to
provide for this demand due to cost and increasing awareness of the environmental impact, and to
recent evidence indicating that further road building may result in increased vehicle kilometres
driven (SACTRA, 1994).

The consequences of congestion on motorways are an increased frequency and severity of flow
breakdown, ie. the formation of shock waves with:
a)     Decreases in operating efficiency,
b)     Higher exposure to accident risk, through increasing speed differentials between vehicles,
c)     Greater environmental dis-benefits.

The range of measures designed to address these problems has increased rapidly over the last 10
years and many promising management techniques are now emerging. However, due to often
inconsistent measures of effectiveness between schemes funded by different bodies, meaningful
comparisons and a clear path and timetable toward the implementation of an integrated interurban
ATT system is difficult. This deliverable will present and quantify where appropriate, 'candidate'
systems based on a range of such approaches.

1.2      APPROACH

To avoid providing a review of ATT systems that is excessively broad at the expense of relevant
depth, activities 110-130 have been undertaken subject to two basic criteria. Firstly primary
consideration has been given to those systems that are about to achieve 'maturity' ie, they are now
at the stage of advanced prototypes or restricted field test in 4th Framework and nationally funded
initiatives. Secondly, in examining relevant developments on other continents, emphasis has been
placed on reviewing developments that have a direct parallel with 'European technology'. (Due to
the fast developing nature of ITS in the U.S.A. and Japan a 'state of the art' watch has been
introduced by initiating contacts with appropriate technical and policy making groups, such as the
U.S. Department of Transportation and specifically the FHWA, thereby allowing additional opinion
to be sought on technical investigations).

Technical considerations (how systems operate, required input/output etc.) have been
supplemented with information relevant to development life cycles, including for example: funding
options to continue development, estimated impact on safety and capacity, estimated cost to end
user, degree of uptake/use and legislative concerns. This information will provide a basis for the
determination of the options to be studied in depth in later project tasks.

Consideration has also been given to evolution that may occur in system operation and use

DIATS                                       Deliverable 1

150796                                                                          DIATS/TRG/WP100

following initial deployment. Here, the consortium has sought to provide an element of foresight
and proposes a range of additional issues that must also be considered in forming plans for the
assessment of these systems. These may include changes to the composition of the driving
population, or to driving behaviour caused perhaps by changing vehicle performances or
adaptations to ATT systems introduced earlier. Thus for example, in order to fully consider the
behaviour of drivers equipped with AICC, one must not only consider how they will adapt to this
particular ATT after its introduction but also how systems such as variable speed limits may cause
them to change from currently observed behaviour before AICC deployment.


The systems considered in this review have been placed in the overall architecture of a potential
intelligent highway system using the CORD (CORD, 1994) functional architecture. (For
completeness, a selection of the full list of CORD functions is presented in Appendix A of this
deliverable). In Figure 1, 'new' systems which may influence motorway capacity are shown in
boxes, and data flows indicated by arrows. Intelligent data processing to highlight potential
problems and propose alternative control strategies such as AID (3.1.2) and ramp metering (3.4.2)
is seen as critical in the process shown. Actuation of these control strategies may then be
performed in general or by additional specialised control implemented for example by imposing
local variations to speed limits (3.1.4). These two control processes (over the main carriageway
and merging traffic) may also be supplemented through independent control of lane allocation (eg.
HOV), 3.4.5. These three elements provide the driver with information on the three main factors
effecting his behavioural decision making process, speed, lane use, and merging. Measures found
in this area of the architectural diagram are termed infrastructure, roadside, or off-vehicle measures.

In the lower part of the diagram, a control loop between the driver and the vehicle actuators driven
by behavioural processes, taking off-vehicle observations of the drivers surrounding as input, may
now be aided by a wide range of in-vehicle co-driver systems that act to enhance driver safety,
comfort or journey time. Some require no input from the driver (eg. 9.4, vision enhancement),
some collect information on the driver and produce feedback (9.2 monitoring driver), while some
act to provide information and if necessary intervene in the vehicle control loop (9.5, Collision risk
estimation - eg. ICC and AICC), with all these systems potentially interacting with the driver
through the intervention of a dialogue management unit (9.7) dedicated to facilitating optimum data

The final potential improvement that may be made to this architecture is to interrupt the link
between external (observable) information and the driver, and formalise an exact link between the
off-vehicle control processes, and the in-vehicle co-driver systems (eg 9.7) in order for the driver to
instantly receive information without the need for periodic signage. Additionally, each vehicle may
transmit information regarding its dynamic state, directly to roadside measurement units (eg. 3.1.1)
to allow control processes to have full information on the state of the road thereby producing an
optimal control strategy. In the most extreme case, this link may be directly established with the
control actuator system, to allow the vehicle path to be optimised off vehicle, with local control
measures such as 9.5, remaining in the loop to control the local dynamics. Such systems, linking
both off- and in-vehicle ATT units are termed 'fully cooperative' or Automated Highway Systems

DIATS                                       Deliverable 1

150796                                      DIATS/TRG/WP100


DIATS                   Deliverable 1

150796                                     DIATS/TRG/WP100


DIATS                      Deliverable 1

150796                                                                         DIATS/TRG/WP100




The main objective of these systems is to replace human control by automated systems, allowing
vehicles to operate at motorway speeds and capacities with regular and stable following distances
of the order of 10 meters. Architecturally cooperative systems involve the most advanced system
design, with on and off-vehicle control technology interacting through two way communication


In Automated Highway Systems vehicles are electronically linked to form so called 'road trains'.
The speeds and accelerations of each vehicle are measured in-vehicle, and then transmitted to
successive vehicles and roadside units for incorporation into control algorithms. The concept of the
Automated Highway System is being developed in many countries (though most notably in the
U.S.A, Japan and Europe). The control architecture for such a system is divided into 4 layers:

i)       The network layer controlling the system-wide flow of traffic. It contains information such
         as the flow on each road, the number of platoons in each, the congestion points, road
         condition and speed limits.
ii)      The link layer contains information such as the number of vehicles in the platoon and the
         desired platoon speed. This layer also handles the protocols for entering and leaving a
iii)     The coordination layer dictates the desired speed of each vehicle in the platoon in order to
         maintain the desired spacing or headway from the preceding vehicle.
iv)      The regulation layer controls the vehicle subsystems (throttle, brakes and steering) in order
         to achieve the desired speed.

Between each layer there is regular two-way communication. The network layer communicates the
desired speed to the link layer, while the link layer sends information about the road condition and
congestion to the network layer. This communication is achieved either through leaky coaxial
cable (LCX) or through cellular communications. The communication between the link layer and
coordination layer comprises the communication between the vehicles of a platoon. In order to
achieve platoon stability, the lead vehicle speed and acceleration are transmitted in addition to the
speed and acceleration of the preceding vehicle. Finally, the regulation layer uses the desired speed
and acceleration provided by the coordination layer in order to compute the desired input for each
of the subsystems.

Most current AHS concepts involve platoons of vehicles with small intervehicular spacing (1-2m)
travelling at highway speeds (100-120Km/h) and large spacing (80m) maintained between platoons,
to ensure sufficient safety (the very small intervehicular spacings allow for low relative impact
speed between vehicles in case of an accident, while the large interplatoon spacing allows for the
platoon the come to a complete stop before colliding with the preceding one).

Current Developments

DIATS                                       Deliverable 1

150796                                                                          DIATS/TRG/WP100

The main developments within Europe in this area have been undertaken in the PRAXITELE
project (Daviet and Parent, and Parent and Fanconnier, 1995), a collaborative effort between
CGEA, EDF, Renault, DASSAULT, INRIA and INRETS in France, which underwent full test for
the first time in mid 1995. The project aims to produce a system that can produce platoon driving
for specialised electric vehicles on urban roads at low speeds (<50 km/h) and at short headways.
Each vehicle is fitted with video cameras and actuators for longitudinal and lateral control, with
special rear markers to aid in the video detection process. Using this system up to 6 vehicles may
be coupled into a platoon (the first vehicle of which is under manual control) with a headway
spacing of about 0.5 seconds. The operation is intended to collect empty vehicles left at arbitrary
locations and redistribute them around the network. The operation includes three phases, the
coupling manoeuvre, platoon driving and leaving a platoon. Current actuators are limited to an
                      2                         2
acceleration of 2 m/s and deceleration at 5 m/s .

Although it performs some of the functions of a fully cooperative system (platooning),
PRAXITELE does not incorporate an exchange of data between vehicles or with roadside
controllers. Thus for fully operational AHS major developments of suitable communication
protocols will be required, as well as in other areas such as the adjustment of controllers to
compensate for the longer time lag present in conventional (petrol and diesel) engines. The most
recent development of note is the CHAUFFEUR project funded by DGXIII as part of the fourth
Framework and led by Daimler Benz. This project, started in early 1996 is to concentrate on short
headway control for HGVs, with intervehicle communication allowing the formation of small linked

Issues and Limitations

Deployment of these systems is generally proposed on segregated lanes, similar to current HOV
(High Occupancy Vehicle) installations because of the additional problems which would be
incurred through using automated vehicles with those driven normally. The costs of the
construction of a network of such facilities will be very high (see Section 2.3.4 for figures).

Early research into cooperative systems indicated that they would provide significantly higher
capacities than conventional lanes, with typical quoted improvements in throughput of over 300%
(eg. Shladover, 1978). However, recent investigations have revealed that the expected increases
may be tempered by a range of other problems that would be generated in the surrounding (non
equipped) traffic (Rao and Varaiya, 1994). For example at a merge, equipped vehicles could be
added to the existing platoons from a parallel merging lane. However, the demand for merging
'places' can soon outstrip the supply provided by passing platoons, restricting capacity to a practical
maximum of 5500 veh/hr (~75% of the theoretical limit for an automated lane). If one considers
this figure on a per lane basis, giving 2700 veh/hr/lane, this produces a maximum flow not much
higher than that currently available. For the above reasons and the large overall cost required for
the successful development of such a system in the short term, it must be viewed as being of
comparatively low relevance to the objective of this project, although some of the technology being
developed in the PRAXITELE project (and in the next few years in CHAUFFEUR) may have
some relevance to the evolution of current Collision Avoidance Systems (CAS) towards low speed
platooning. This topic is further discussed in Section 2.2.1.


DIATS                                       Deliverable 1

150796                                                                         DIATS/TRG/WP100

These systems comprise those in the lower half of the architectural plan shown in Figure 1, and are
dedicated to enhancing the sensory performance of the driver and producing a more optimum

2.2.1    Collision Avoidance Systems (CAS)


These devices seek to control the speed of the vehicle in order to maintain a 'safe' distance between
consecutive vehicles in the traffic stream, subject to the condition that the driver may intervene at
any time and is ultimately responsible for the behaviour of the vehicle.


The main goals of CAS (more generally called AICC systems (Autonomous Intelligent Cruise
Control)) are to add a function of automatic or advised distance keeping to the conventional cruise
controller. This requires distance measuring equipment (usually infra-red laser, or radar rangefinder
units), onboard computation abilities ("Intelligence") and actuators for (semi-) automatic control
modes. Unlike Cooperative systems, the driver may override the control systems and is more
clearly responsible for all manoeuvres. Following a large number of field tests over the last 5-7
years these systems are now on the verge of market introduction.

Several differing systems exist under the general classification of CAS, and differing abbreviations
are in use wide use (typically there is a confusion between ICC and AICC). However, the
following four basic types can be identified, all of which feature a target (safe) distance calculated
from a constant headway which ranges from 1.2 to 2.0 seconds and are aimed at increasing a
driver's comfort and rear end collision avoidance capabilities:-

-        Collision Warning systems (CW), where the driver gets warning information through an
         MMI but stays fully within the control loop.
-        Assisting systems, a tangible signal is given when the headway becomes too short, eg. in
         the form of increased accelerator pedal resistance, termed Haptic Feedback (HF).
-        Semi-automatic driving or Intelligent Cruise Control (ICC), where control is exerted on the
         throttle in order to obtain a set headway. (The driver may still accelerate the vehicle past
         this recommended time headway by applying pressure to the accelerator). The following
         distance may either be pre-set by the manufacturer or by the user according to driving style.
-        Automatic driving (Autonomous ICC or AICC), where control is exerted on both the
         throttle and the brakes in order to attain the set headway

Current Developments

i)       Within PROMETHEUS, Hella in Germany (Hellmuller, 1995) have developed an
         intelligent speed controller, that measures intervehicle distance by a pulsed infra-red beam
         and also by radar. The target headway in following mode is set at 1.8 seconds, and initial
         prototypes were first tested during 1995. Introduction to the market is expected within 3
         years and future developments look to expand the system to a true AICC, in which the
         vehicles brakes may be used to enable a higher rate of deceleration. Purchase and

DIATS                                       Deliverable 1

150796                                                                            DIATS/TRG/WP100

         installation costs have been estimated at 600-1000 ECU.

ii)      BMW AG have developed a prototype HF system which was tested in mid 1994 and uses a
         3 beam infra red laser. This system is slightly different from most of this kind, in that it
         allows the driver to set his desired headway between 1 and 2 seconds (Naab and Reichert,
         1994). The control strategy has proved satisfactory in most tests and current efforts are
         focusing on the development of a fully automatic version where the system can also control
         the brakes and in solving problems of picking up the wrong target vehicle when rounding
         bends. Market introduction is expected within the next 3 years.

iii)     Matra and Renault, have developed a CW device, which has been tested within the
         DRIVE2 TESCO project. Tests were performed using a series of densely spaced roadside
         beacons to allow for precise recording of all data within the platoon of test vehicles. When
         passing a beacon a vehicle delivered information about its own velocity and acceleration
         and receives its actual position as well as the position, speed and acceleration of the
         preceding vehicle. Within the test system itself, two risk levels are defined and are
         displayed to the driver when the thresholds are exceeded. Both risk thresholds depend on
         the distance, speed and acceleration. The lower level displays two arrows (chevrons) and a
         relatively small vehicle on a screen, while the upper level shows one arrow and a large
         vehicle. The display option is based on a calculated risk level within the range 0-100,
         according to two different procedures: an algorithm developed within PROMETHEUS
         based on a ratio of 'relative safe distance' and the actual distance, and the second, an
         algorithm taking the required deceleration for a braking manoeuvre as a criterion (Sala and
         Pressi, 1994).

         In the Matra approach a monochrome screen is placed on the instrument panel while with
         the Renault system, a colour screen is positioned in the centre on top of the instrument
         panel. (The system proved effective in field tests, although it was noted that the data
         exchange with a beacon could be interrupted by another vehicle in the line of sight.
         However, as the measurement configuration using roadside beacons was only a special
         solution for a field test there are no implications for the introduction of this warning system
         which will be based on conventional distance sensors).

iv)      Opel and Daimler Benz in Germany have both developed true AICC units (where up to
         25% of the capacity of the braking system may be applied) through PROMETHEUS,
         which were first tested in mid 1994. In this system, distance and relative speed are
         measured by a radar with a 12 degree coverage. The driver is able to select a headway of
         1.2, 1.4 or 1.8 seconds. The system can be deactivated by two traffic related events, the
         first is the use of the brakes by the driver, and the second is if the vehicle speed falls below
         40 km/h (a warning tone sounds to indicate deactivation in this case). The overall control
         strategy has generally proved to be effective, although some difficulties have been apparent
         in narrow curves and in strong rain. The price of the system is estimated to lie in the range
         of 600-1000 ECU.

v)       Jaguar and LUCAS in the U.K have developed two systems (Tribe et. al., 1995). One of
         these is a joint CW and full CAS system, where an HF function warns the driver when he
         attempts to drive too close to the vehicle in front. Should a second closer headway be
         exceeded, the codriver unit will take control and decelerate the vehicle (using both

DIATS                                        Deliverable 1

150796                                                                         DIATS/TRG/WP100

         accelerator and brakes) in order to maintain the minimum safe headway, accompanied by
         an audible warning. The second system is a fully functional AICC unit which acts to both
         accelerate and decelerate the vehicle as appropriate (up to a maximum of -3 m/s ), similar in
         nature to those developed by Opel.

vi)      Rover in the U.K. have developed a CW system that has seen extensive investigation within
         several DRIVE2 projects (see below). This is based on a multi beam radar unit and
         warnings given to the driver through the use of an HF throttle based on a calculated time to
         collision threshold.

In addition to these specific development programmes, several 'horizontal' activities have been
undertaken in the last 4 years to compare the MMI and user acceptability aspects of the differing
systems. The most general of these (focusing on MMI aspects) have been the range of projects
undertaken as part of the Area 4 workplan of DRIVE2, which, in addition to the TESCO project
described above has included the projects:-

*        EMMIS. Dedicated to the evaluation of man machine interfaces, primarily through the use
         of driver simulators. Although not specifically aimed at the evaluation of CAS systems
         (navigation and medium range pre-information functions were also investigated), visual and
         acoustic warnings for CW were investigated, as was a fully operating AICC system.
         Vehicle manufacturers participating in this project included BMW and Fiat.

*        HARDIE. Dedicated to the evaluation of the differing modes through which
         communication could be performed with the driver for CAS, which undertook tests on a
         range of real roads in the U.K., primarily focusing on the use of HF and acoustic (both
         single 'buzzers' and spoken word) warnings in CW (both in isolation and in combination)
         based on a simple 1 second recommended headway.

*        DETER undertook real world trials in addition to simulator tests on the effects of HF and
         verbal messages on driver compliance with a 1 second headway regulation, which also
         included experiments on automatic enforcement systems (see section 2.2.3) involving a
         consortium which included Renault.

*        HOPES undertook a horizontal safety evaluation of a range of ATT measures, including a
         range of cooperative driving options, using driver simulators. The CAS systems evaluated
         included the CW system tested in the TESCO project, the AICC system developed by a
         range of manufacturers within PROMETHUES, and the HF approach developed by Rover
         within the ARIADNE project.

The findings of the above projects, which have used a wide range of differing methods and
evaluation procedures, have been summarised in the following two studies:

DIATS                                       Deliverable 1

150796                                                                         DIATS/TRG/WP100

i)       The Driver Assistance and Cooperative Driving Cross Project Collaborative Study
         (DACD), conducted as part of the CORD project within DRIVE2, has attempted to form a
         consensus on the effects of, and driver attitudes to, the systems under test through 'meta
         analysis' (CORD, 1996). Based on results from the DRIVE2 projects mentioned above,
         the main conclusions of the study were that CW and HF systems acted to increase average
         headways, while ICC and AICC acted to reduce them. It was notable however, that the
         magnitude of these changes were highly dependant on the mode of investigation (simulator
         vs real road tests). However, one of the most important findings were those from 'soft data'
         sources such as driver acceptance questionnaires, which indicated that all the CAS systems
         were viewed as being beneficial and not particularly demanding in terms of workload.
         Some systems were viewed as annoying, in particular verbal (and in some cases visual) CW
         systems. ICC and HF systems were viewed as being relatively easy to use and a distinct aid
         to the driving process.

ii)      Work conducted within PROMETHEUS (Becker et. al., 1994 and 1995) has established
         several more substantial trends to be considered in the development of CAS systems.
         Findings were based on extended public trials using prototypes from all the major German
         manufacturers. These investigations focused on three major issues:

         -      Technical layout and MMI, incorporating drivers comprehension of the operation
                and limits of the device, the acceptability of the operational strategy, needs for the
                inclusion of autonomous braking and the definition of an optimal MMI.
         -      Driver behaviour, including changes in mental workload, delegation of
                responsibility to the system, reduction in attention and changes in driving style
                regarding safety.
         -      Product evaluation and acceptance including users' perception and effect of market

The nett impact of these measures on the traffic flow as a whole has also been evaluated within
PROMETHEUS (PROGEN) and DRIVE programmes, which has indicated that in the case of
AICC, it may be possible to harmonise speeds (ICARUS, 1992) by placing 70% of the vehicles to
within 6 kph of the average speed (as opposed to the current 25%), increase the maximum flow
before the onset of flow breakdown by up to 13% (Broqua, 1992), decrease the frequency of un-
comfortable braking events by 79% (Benz, 1994), and potentially reduce fuel consumption by 2.5%
(Zhang, 1992).

At the end of PROMETHEUS and DRIVE2 in 1994, the opportunity for continuing research with
these systems was severely reduced, with no general replacement programme for PROMETHUES
being available. Work has since continued in a number of 4th Framework projects funded by
DGXIII (the most notable of which being AC-ASSIST/ROADSTER, incorporating Jaguar, Rover,
Renault and Volvo and continuing work on 'mainstream' AICC applications), and under a range of
national initiatives forming during 96, (eg PROMOTE in the U.K, MOTIV in Germany and
PREDIT in France).

DIATS                                       Deliverable 1

150796                                                                            DIATS/TRG/WP100

Issues and Limitations

With the breadth of systems now under test and further development it seems likely that a degree of
commonality may soon form between system manufacturers, and that only two main types of
system will effectively reach market. One such system is AICC, where those manufacturers who
currently have only developed ICC units are actively pursuing this as a next stage (perhaps partially
driven by the observation that drivers have expressed a degree of trepidation with ICC (Becker et.
al., 1995) over when they are required to intervene in a harsh deceleration situation, expressing a
desire that such systems should be supplemented by an emergency brake). Another system which
may see market is the 'safer' CW (and HF) units, which may either be used as an 'introductory level'
to CAS or as a single product should manufacturer policy decide not to fit AICC units through
concerns over user mis-adaptation to the system. Despite the growing body of both technical and
behavioural research that has been undertaken, there are still several unresolved problems:

i)       The effect of CAS systems on the following headway has been reported as both decreasing
         and increasing following headways. This anomaly may in part be due to the method of
         assessment where it is not yet clear how, or if, driver behaviour changes when assessed in a
         simulator. This may however also be due to drivers interrupting the system at high
         densities, where the continual acceleration-deceleration phases of the car have been
         reported as being distracting (Tribe et. al., 1995), leading the driver to turn off the system
         and drive manually.

ii)      Concern has been expressed over the performance of the AICC when a vehicle moves in in-
         front of the equipped vehicle as in a lane change, with the control system over reacting and
         deceleration too harshly. In addition to reducing comfort, this behaviour would also have
         the effect of reducing the stability of a stream of vehicles.

iii)     The degree of uptake of CAS may be effected by the degree of effectiveness of off-vehicle
         systems, where the potential success of integrated systems combining AID, ramp metering
         and variable speed limits for example, may contrive to give drivers a greater expectation of
         the CAS system in terms of its personal benefit.

iv)      Additionally, behavioural changes that may result from 'traffic calming' through the
         enforcement of on line speed control systems, may reduce the impact of CAS on the traffic
         flow as a whole and require a greater penetration level to be achieved before any overall
         (non-personal) benefits can be achieved).

v)       CAS systems may actually decrease traffic safety in the long term, in that once a user is
         familiar with the system he may come to rely on it too much (particularly ICC), at times
         assuming it to be active in situations where it may not be so, such as urban driving.

vi)      The ability of a large penetration of these systems to substantially homogenise flow may
         produce problems at intersections, where merging vehicles may find it increasingly difficult
         to locate and move into, a suitably large gap. At high merging flows this could result in on
         ramp queuing or short headway (and hence high risk) lane changes with additional
         implications to (ii) above.

viii)    In 'bad weather conditions' it is generally perceived that the ability of the vehicle sensors to

DIATS                                        Deliverable 1

150796                                                                         DIATS/TRG/WP100

         accurately detect the next vehicle may make driving safer. However, it is possible that such
         a reliance may increase average driver speeds hence counteracting any advantage gained.
         Also, almost all of the sensors have a reduced operational range and accuracy (typically
         from 120m down to 70m or so) in these weather conditions, with laser sensors suffering
         particularly badly.

Research work currently being undertaken may soon be able to minimise if not totally eliminate
some of these problems. Firstly, increasingly complex sensor arrangements and processing
software may soon enable subtle driving problems such as those in (i) and (ii) to be identified and
adjusted to an increasingly 'natural' manner of operation, and in the case of (vi), a limit being
applied to the maximum controlled speed to account for the sensor inaccuracy. Enhancement of
the current CAS sales package with other systems, such as dialogue mangers and driving
impairment monitoring systems, will undoubtedly increase the attractiveness of the use of the AICC
or CW and to some extent counteract the problems encountered in (iii) and (iv). The potential over
reliance on CAS (v) is currently being investigated as part of the DGXIII UDC project to explore
the implications and opportunities for the deployment of CAS in urban regions.

2.2.2    Vision Enhancement


To improve the driver's vision of the road environment in poor visibility and dark conditions
without the use of dazzling full beam headlights.


Research in this area has focused on the development of three types of system. Firstly, Ultraviolet
(UV) based headlights have been developed. This technique involves using full beam UV
headlamps alongside ordinary dipped headlights. If road users and infrastructure are dressed in
materials which fluoresce under UV light then full beam visibility can be achieved. UV light has the
advantage that it is not visible and so cannot dazzle. This system is relatively cheap to install on
vehicles, but would require expensive infrastructure investment.

The second, so called, Near Infra-Red (NIR) system operates in the infra-red region near to visible
light. This technique requires a full beam infra-red headlamp alongside the ordinary dipped
headlights. Infrared light is not visible and so cannot dazzle. The reflected infra-red light is then
observed by an infra-red sensitive camera. The infrared headlamp can be fitted with a 45 degree
polariser to prevent dazzling the cameras of oncoming vehicles. The signal obtained may be
processed and presented to the driver by the use of a head-up overlay display which superimposes
the information on a normal driver's view. A pulse intensified version can be used to increase vision
in fog and rain where back-scatter would normally reduce the effectiveness of this technique.

The last approach uses Thermal Imaging or Far Infra-Red (FIR). This is a passive technique which
does not require any special illumination. The infra-red camera fitted to the front of the car is
sensitive to the infra-red energy emitted by objects of differing temperatures. This energy occurs in
the infra-red region quite far from visible light. The information from the camera may be processed
and displayed to the driver in the same way as in NIR systems using a head up overlay display.
Pedestrians and cyclists show up particularly well with this technique. The contrast is not affected

DIATS                                       Deliverable 1

150796                                                                         DIATS/TRG/WP100

by the ambient lighting conditions. The sensors are expensive and require cooling, but research is
taking place into cheaper versions. This system can provide better visibility than main beam visible
light headlights. No companies were found who were currently developing this technology with
the aim of using it in vehicles.

Current developments

i)       Fiat in Italy, are currently investigating pulse intensified NIR systems as well as a similar
         system which works in the visible light region and that employs associated image
         processing algorithms. Particular attention is being given to the experimental techniques
         involved with research into such systems including fog generation and performance
         assessment (Panizza and Levizzari, 1990).
ii)      Jaguar together with GEC Marconi and Pilkington in the U.K. have developed a
         demonstrator vehicle which incorporates an NIR system and effective head up display
         (Evans, 1995). Current research is focusing on reducing the (currently high) cost of an
         effective head up display and investigating human factors issues.
iii)     Renault in France have assembled a demonstrator vehicle incorporating an NIR system.
         The system is undergoing fine tuning particularly regarding the operation of the MMI.
         Concern has also been expressed about possible side effects of the system and product
         liability (Augello, 1993).
iv)      Volvo have primarily been active in investigating ultraviolet light headlamps with Ultralux
         (a company owned by Volvo, Saab-Scania, and others), and now have technology ready
         for development. Sweden's National Roads Administration has prepared 100 km of
         fluorescent road infrastructure in Gothenburg and Trollhattan for forthcoming tests.

The main body of work so far that has attempted to assess driver response to the systems has been
the DRIVE2 Project EDDIT (Stahl et. al. 1994). Although this work focused purely on elderly
drivers, results of tests using both NIR and UV technologies showed an increased sight distance,
and a subjective increase in the drivers assessment of confidence and safety.

There has so far been no assessment of drivers' desires to purchase such units, (or indeed any
indication given of likely cost and 'time to market'). However this is likely to be highly dependant
on the type of system. The vehicle installation cost of UV systems is likely to be low, but the
infrastructure costs high, while for IR systems, there is no infrastructure cost, but the in-vehicle
costs may be high.

Issues and Limitations

These systems should theoretically reduce accident rates with drivers able to interact with
neighbouring vehicles over a longer range, and hence may operate safely at higher speeds.
However, accident rates may increase due to the higher speeds, and drivers not fully
accommodating for the behaviour of other drivers in non-equipped vehicles in situations such as
lane changing for example.

The overall impact of this type of system, is likely to be relatively minor in recurrent congestion
scenarios, however it is likely to have greatest effect in minimising the frequency of non recurrent
congestion scenarios resulting from incidents, where the system theoretically reduces the frequency
of incidents. This is a scenario which is very difficult to investigate through the use of simulation

DIATS                                       Deliverable 1

150796                                                                          DIATS/TRG/WP100


2.2.3     Driver Status Monitoring and Support (incorporating Automatic Enforcement)


To monitor the driver's abilities to affect the behaviour of the vehicle, as a basis for predicting the
characteristics of the vehicle's trajectory and providing suitable information.


The driver, except in circumstances where fully cooperative systems have been installed, is the link
between the collection and provision of information and the behaviour of the vehicle. Monitoring
of driver's status assists in estimating the drivers' reaction to information and effectiveness in
conducting the driving task.

Such systems have primarily been examined in recent years within the DRIVE2 DETER project
(Brookhuis, 1993), itself a successor to the DRIVE1, DREAM project (Brookhuis et. al., 1991).
The system developed information on driver status primarily for on-line driver tutoring. Particular
emphasis was placed on measuring and validating a range of primary indicators of driver alertness,
to specify and produce a vehicle Driver Impairment Monitoring system (DIM), which is based on
observing speed and headway violations, as well as several factors related to compliance with road
signs. The information required on the locally acceptable values of these variables is obtained from
either on or off-vehicle sensors (eg. roadside beacons or AICC). Once this information is available,
an 'expert system', based on a neural-network approach, compares observed behaviour with
historical data on a drivers style, and decides whether to use instructional support messages or to
start enforcement, via logging a violation, or even activation of hazard lights and slowing down the
vehicle (Fairclough et. al., 1994).

Driver status monitoring has also been investigated to some extent in the DRIVE2, ARIADNE
project (Groeger et. al., 1994), although here, data on driver status was used primarily as a
predictive tool in the establishment of a dialogue controller (see later). Driver feedback messages
were seen as providing an aid to improvement in driver behaviour, both in the short term and in the
long term. In the former, if there has been deviation from a set of rules for acceptable driving, for
example speeding or lane departure, an appropriate message is sent via a scheduling module to be
presented to the driver. In the long term, details of deviations from the rules are passed to a
History and Instructional Module (HIM). In this case, a long term improvement in driver
performance can be gained by recording and monitoring abnormal driving behaviour, output either
on demand or automatically to the driver if some predetermined criteria are reached. This approach
can be further augmented by programming the system to give prospective messages when the
system thinks a scenario is going to occur that has been the subject of warnings in the past.

DIATS                                       Deliverable 1

150796                                                                          DIATS/TRG/WP100

Additional work has also been performed in the DRIVE2 project SAMOVAR (Fincham et. al.,
1994, 1996), building on technology developed in the DRACO project regarding an accident data
recorder (ADR) which may be used to record information on driver behaviour and violations. The
systems developed have undergone several extensive field tests, primarily in fleet vehicles, in several
DRIVE2 city projects, and provided data on potential changes in driver behaviour from vehicles
fitted with the ADRs. Although it has been found that the fitting of such units can reduce accident
frequency by approximately 28%, it is important to note that this has been obtained in urban
conditions. No figures are available about the effects that ADRs may have on motorway driving
(SAMOVAR, 1995).

Current Developments

Work performed in DETER is now being continued in the SAVE project, with continued industrial
involvement from Renault and Phillips and new participation from Fiat. The project now focuses
solely on the development of driver impairment detection within the vehicle (now also including the
effects of drugs and alcohol), and how that information may be used by a codriver system, to slow
the vehicle and activate emergency help systems. Extensive tests using a number of instrumented
vehicles and simulators, and on road tests at several differing pilot sites are expected to be

The SAMOVAR project is currently inactive and no plans are available for its continuation. This
places the future of large scale manufacture of such units in jeopardy, because it may soon becomes
uneconomic to continue the manufacture of a small number of units. Current costs are
unacceptably high (for the individual) at 700 ECU, although for a fleet user these costs are easily
offset by potential savings on vehicle costs through changes in driving style. It is hoped that costs
could be reduced to nearer 200 ECU if full-scale production is ever undertaken.

Issues and Limitations

Most drivers would not willingly choose to purchase a vehicle with a driver monitoring unit,
despite the general success in their development (a comparatively low false alarm rate of 21% was
attained) and tested subjects showing a high acceptance of the systems. Also, enforcing the use of
such units in order to reduce accidents through monitoring for drunk driving or drowsiness could
infringe civil rights. Therefore, potentially major markets would be vehicle fleets, or compulsory
use by offenders, particularly those found guilty of drinking offenses. Similarly the use of ADR's is
likely to be severely limited through public perception, where doubts over reliability and accuracy
(which seem to be unfounded), civil rights issues, availability of data etc, and cost would seem to
again indicate that the system would best be received in commercial fleets.

The overall benefit of these two methods to the overall operation of the motorway is therefore (like
vision enhancement) likely to be small, and would most likely only act to reduce the number of non
recurrent congestion scenarios caused by accidents. However, the impact of ADR's on the
behaviour of commercial fleets should be considered, as this may have behavioural implications.

DIATS                                       Deliverable 1

150796                                                                         DIATS/TRG/WP100

2.2.4    Dialogue Management


Seeks to optimise the transfer of information from any co driver system to the human driver, via
adaptation of the MMI options available, prioritising and allowing sufficient time for the driver to
receive, understand and react to system messages.


Systems of this kind are concerned with determining (and managing) the requirements for the
operation of intelligent codriver systems to be consistent with the information requirements and
performance capabilities of the driver. The tool developed, commonly termed the 'dialogue
controller', acts to rank messages for presentation to the driver, and determine the optimum
selection of media (eg. visual, audible, tactile) and the timing between the messages.

Basic systems effectively decide which of any potential messages provided by, for example,
navigational, radio or collision avoidance/warning systems should be given first (generally the one
with the most important message as regards safety takes precedence). A more intelligent version of
the prototype, incorporating an 'intelligent adaptive presentation mode' also enables the lower
priority message to be postponed, by estimating the workload the driver is likely to be under in the
next few minutes and deciding when he will be able to cope with the second message. This is
accomplished by cross checking the traffic situation (as regards complexity) with driver experience
on a series of look-up tables. The road situation is ascertained by a fusion of available sensor data,
for example, range to the next vehicle from an AICC unit indicating traffic density, and/or
navigational information from GPS units/roadside beacons, regarding proximity and nature of
junctions (Verwey et. al., 1994).

Once this 'maximum input' that the driver can take has been established, the controller may decide
how to transmit the information (visual vs spoken message or tones vs active accelerator pedal), the
time to leave between messages, how long to display visual messages, to avoid overloading the
driver, distracting him and creating a safety problem.

Current Developments

Research into this area has been focused in recent years by the DRIVE2, ARIADNE project, an
extension of the work performed in the DRIVE1 project GIDS, and has produced a prototype
dialogue controller (in addition to contributions to Driver Monitoring systems). Throughout this
project a great deal of experiments have been conducted to establish the 'look-up' tables, and the
appropriate combinations of output according to a range of differing road situations and driver age
(eg. Janssen et. al., 1994). Since then however, no major 4th Framework projects have so far been
funded into this area and it is believed that individual manufacturers will continue this work under
internal funding (Volvo and BMW) and under national initiatives (Rover in the UK, as part of the
RTA project). It is likely however, that this work will now focus on the integration and
development of ARIADNE findings into marketable systems, effectively moving away from the
earlier more speculative 'generic' investigations.

Issues and Limitations

DIATS                                       Deliverable 1

150796                                                                            DIATS/TRG/WP100

Because this type of system is not a separate 'add-on' option for vehicle buyers, but rather is an
integral part of any new in-vehicle ATT system, it is unlikely that it will affect the cost of the unit it
is intended to support, but will act as an increased 'selling point' making systems 'easier to use'
(McMurran, 1994). This may itself have some implications for the final acceptance and degree of
penetration of CAS systems, but is unlikely to have any impact of its own that can be easily
investigated without the need for extensive further research on behavioural modification. (For
example, will drivers learn to trust the system and not become hesitant, believing that the system is
erroneously holding back important information. If this happens with CW for example, drivers may
'hang back' with increased headways and resultant implications to capacity).


These systems act to effect the behaviour of drivers as a whole, and comprise systems in the upper
half of the architecture shown in Figure 1. Whereas in-vehicle systems perform continuously to aid
drivers, these (at present) may only act intermittently, with information transferal being governed by
the spatial frequency of signs used to convey messages, and the time frequency over which a driver
can smoothly react.

2.3.1    On-line speed control Systems


To control the speed of vehicles by changing local speed limits as posted on VMS gantries.


Experiments with variable speed control were initially performed in the 1960's and 70's in many
locations (Morin, 1982), but the majority of the automatic systems of concern to this project have
been developed since about 1989. The main objectives are to reduce traffic speeds under heavy
flow, or poor environmental conditions in order to warn drivers in advance of congestion or
queuing ahead. In some cases, speed control has been specifically introduced to reduce accidents in
fog. In other cases, it is hoped that reducing the variability in speeds between vehicles may prevent
flow breakdown and lead to lower overall journey times.

The systems consist of devices for measuring traffic conditions, gantries for displaying the speed
limits and a control algorithm for determining the settings from the data collected. The algorithm
for calculation of the speed level can be simple or complex and can be based on flow (in veh/h or
pcu/h), speed and density (or occupancy), as measured across all carriageway lanes or individual
lanes, collected via induction loop detectors in the road surface or video cameras on overbridges.
The systems usually have rules for longitudinal smoothing (to ensure consistency in speed limits as
drivers travel along the section), a smoothing mechanism for the traffic data so that sudden changes
do not cause oscillations of speed limits in time, and hysteresis, so that speed limits are not raised
until it is certain that normal conditions have resumed.

Across the EU, these systems have been implemented according to differing goals. In Germany,
the main benefits are seen as a reduction in accidents; in the UK it is hoped it will lead to lower

DIATS                                        Deliverable 1

150796                                                                         DIATS/TRG/WP100

journey times. Other possible benefits are increased flow levels, more predictable and more
consistent journey times, reduced driver stress and reductions in vehicle noise and air pollution.
Systems currently in use are generally reliable and perform satisfactorily, and have fail-safe
mechanisms built into them. Most schemes involve equipping 10 to 20 km of motorway (both
directions) and installation costs are around 1 Million ECU/km.

Current Development

i)       In the U.K., a 23 km section of the M25 London orbital motorway has had speed control
         since August 1995, with speed limits being reduced to 60 and 50 mph (97 and 80 km/h).
         The system used fixed time plans initially, but since September 1995 a flow responsive
         system has been in operation developed by TRL (Hardman, 1996). A more complex
         algorithm involving both speed and flow will be introduced in late 1996. Enforcement of
         the system is high, with speed cameras in operation along all of the equipped stretch. No
         results are yet available as to the effectiveness of the system and the degree of compliance,
         although an assessment of the pilot scheme will be made in 1996.

ii)      In Germany, about 40 motorway sections (totalling over 200 km) have traffic responsive
         speed control with limits of 120, 100, 80 and 60 km/h, under the control of the motorway
         traffic control centres in different Lander. Compliance with the systems is generally high
         (some systems have automatic enforcement, others none). All are very complex and
         introduce speed limits under heavy traffic conditions, when variance of speed of the left
         lane exceeds a certain limit during medium or high traffic load, when fog or rain are
         detected, in addition to changes in speed, flow and density compared with threshold values
         as used elsewhere. Studies have shown a reduction in the injury accident rate per vehicle
         km of between 14 and 37 per cent, relative to other motorways without speed control. The
         majority of these systems have been developed by Heusch Boesefeldt and some are
         currently being extended. Some work is being carried out on new algorithms like the
         Ferrari method (SPECTRUM, 1993) or Kalman filtering, both aiming at an earlier
         congestion recognition. Other approaches use an improved traffic flow model for density
         determination (Kerner and Konhauser, 1994) or are working with congestion recognition
         from video image processing (e.g. IMPACT). Another direction of research is fuzzy
         control application. All these activities are aimed at sophisticated enhancement of
         algorithms already in use.

iii)     In the Netherlands, a system of 70 and 90 km/h limits (non enforced) was introduced in
         1992 on a 20 km length of the A2 Amsterdam-Utrecht, based on measurements of speed
         and flow. Research (Smulders, 1992, Hoogen and Smulders, 1994) showed that traffic
         speeds were reduced along with the number and severity of shockwaves, but no positive
         effects in terms of improved capacity were observed. It is believed that a new series of
         trials may soon be undertaken, this time with enforced speed limits.

DIATS                                       Deliverable 1

150796                                                                            DIATS/TRG/WP100

iv)      In Belgium, a system on an 8 km section of the E17 Ghent-Antwerpen on the approach to
         a tunnel is currently being developed, with traffic data being obtained by using the CCATS
         image processing system linked to video cameras on overbridges. An algorithm using
         traffic occupancy and speed will determine the speed limit setting (TRITEL, 1994). (In a
         related, though non-automatic, system in current use on the E40 Gent-Oostende, police
         patrol cars slow traffic down in order to harmonise speed).

v)       In Denmark, two systems are under development (Danish Ministry of Transport, 1995).
         On the approach to a tunnel in Aalborg, speed limits of 50, 70 and 90 km/h will be set
         according to the measured traffic speed. The other system is in the planning stage and is to
         be based on the motorway network around Copenhagen.

Issues and Limitations

Although no comparative work has yet been undertaken across the growing number of test sites in
Europe, it seems likely that the effectiveness of the approach is highly dependant on the level of

A factor in the degree of flexibility available in the application of a suitable control system, is the
spatial and time intervals over which data is available for calculation and with which the limit may
be changed. For example, in order to exact a finer degree of control (and hence ensure that a
greater number of drivers are exposed to the 'ideal' speed), the spacing between signing points and
the recalculation of the value must be minimised to avoid large numbers of vehicle passing and
using an inappropriate limit. However, the spatial frequency of the signing points clearly has a
minimum and in practice this is determined by the financial outlay available, or the availability of
existing features. Changing the limits too frequently in time is also difficult as the driver must not
be presented with a constantly fluctuating limit which is difficult to consistently observe and
additionally difficult to both practically and legally enforce. A full analysis of this system (and hence
further development) in also hindered by a lack of information on driver reaction.

Despite some minor setbacks, this approach now looks likely to become widespread at most
congestion black-spots across Europe over the next 5 years, where it will form part of a general
intelligent motorway system which may include route diversion, incident management and AID,
queue detection, roadworks operation.

Although, no major research initiatives are currently underway in this area there is great potential
for future improvements to be made, in particular by considering the effects of HGVs and lane by
lane control to enable a finer degree of control to be available, and in integration with junction
management measures such as ramp metering.

DIATS                                        Deliverable 1

150796                                                                           DIATS/TRG/WP100

2.3.2    AID (Automatic Incident Detection)


To detect incidents and congestion, in order to allow relevant intervention, information
dissemination and control, to reduce the impacts of the incidents on the traffic flow and safety.


Any unexpected event that decreases traffic flow or safety levels may be considered as an incident.
The direct cause of these events cannot usually be directly detected but they are apparent from their
effects on traffic flow parameters such as volume, density and speed. Incidents may be detected
when for example, a vehicle is seen to stop for more than a certain number of seconds, or if queues
form and speed decreases below a preset threshold. These events can be detected using real time
and historic data which forms the basis of AID systems.

AID systems are typically assessed according to criteria such as rapidity of detection, the detection
rate (= number of alarms / total number of real known events), the false alarm rate (= number of FA
/ number of alarms recorded), the frequency of false alarms (= number of FA / 24 h / detector) and
the identification of features of the incident (type, duration, severity). Several of these measures are
competitive and a compromise must often be reached, between a high detection rate (which would
decrease the threshold of the detection), and a low false alarm rate (which would increase the
threshold of the detection). Current approaches are generally reliable in moderate to high demand
situations and all systems have been subject to a detailed calibration process (eg. Stephanedes,
1993), with some approaches switching between several available algorithms according to
conditions (Cohen, 1993). Another important factor in characterising systems is in the type and
configuration of detectors used (eg, induction loop, video, infrared or microwave, see Busch,
1990), typical among which are:-

*        Pair or multiple spot detectors (loop detectors or beacon on the roadside) to provide flows,
         speeds, (and occupancy in the case of loop detectors). The information is processed every
         second, or integrated on a longer period (30 seconds or more). The information of the first
         detector is shifted by a time lag corresponding to the driving between these two detectors,
         with the algorithm comparing the two responses eg. HERMES, IDRIS.

*        Single spatial detectors (radar, video), providing the positions and speeds of vehicles, eg.
         BEATRICS (radar), INVAID/ Motorway (using the TRISTAR video detector). By
         combining a number of spatial detectors, a stretch of the highway can be monitored.

A complete review of European urban and inter-urban AID systems has recently been performed
within the DG XIII CORD project (CORD, 1995) and includes detailed information concerning all
(currently over 10) systems in active use within the CEC, a representative selection of which are
presented below.

Current Developments

i)       Siemens AG, have developed several methods, among them AID using Kalman filtering,
         which has been developed in association with Technical University of Hamburg (Cremer,

DIATS                                        Deliverable 1

150796                                                                         DIATS/TRG/WP100

         1981) under partial EU funding, and was first tested successfully in 1994 within the
         HERMES project (Cremer and Morello et. al., 1994). The system relies on loop detectors
         for volume and speed measurements at points from 1 to 5 km apart, and sampled at time
         steps between 10 sec and 1 minute. For detector distances from 2 to 4 km the method has
         proven to be superior to any other comparable detector based methods. Reaction time is in
         the range of 2 minutes and high detection rates and low false alarm rates are reported but
         not yet quantified (Busch et. al. 1994).

         A similar system which is also based on the Kalman-Filter approach has been developed
         and successfully tested in the field by Heusch-Boesefeldt. This system is especially
         designed to cope with longer distances of 5km to 10km between two detectors and to
         support automatic message generation by RDS/TMC information systems.

ii)      The INVAID system is primarily based on the use of video images, and has been developed
         by INRETS/DART both internally, and as part of several DRIVE projects, along with
         VELEC, using the TRISTAR detector unit, taking images collected from video cameras
         approximately every 500m (eg. INVAIDII, 1994, 1995). The system was initially
         developed in 1985 and has been under test since 1989, on a number of French motorways
         including the A43, A8, A7 and on roads under the jurisdiction of COFIROUTE. The
         systems is able to detect stopped vehicles 80% to 90% of the time with a delay of
         approximately 15 seconds, however some degradation is noticeable (in common with most
         video based methods) when shadows occur, and becomes exceptionally difficult in fog.
         Congestion is generally detected with a frequency of 90%+ with a delay time of a few
         minutes, and a false alarm rate of 0.3/24h /km. Estimated costs of the installation of the
         systems currently run to 120 kECU/km, with a yearly maintenance cost of about 5% of the
         initial outlay (Morin, 1996).

iii)     The BEATRICS system is based on radar and has been developed by the Thomson
         company, using microwave sensors placed on poles on central reserve of the road (Roussel
         et. al., 1995). The individual speeds of the vehicles passing in the field of the sensor are
         directly measured, estimations of their acceleration derived, and the traffic volume counted.
          The results (statistics, alarms) are broadcast to a central unit by a transmission network.
         Two measurements principles have been developed, a radar covering an area of 100 meters
         long and 5 lanes in each direction and a radar, covering up to 4 lanes in each direction at a
         single point. This system was successfully tested between 1992 and 1995 on the A46N, as
         part of the DRIVE2 MELYSSA project, and current work is focusing on implementation
         near toll booth areas, extensions to differing traffic conditions, and improvement of the
         processing and of the use of measured traffic data. The system has been quoted as
         achieving a detection rate of 87%.

DIATS                                       Deliverable 1

150796                                                                         DIATS/TRG/WP100

iv)      Recently, the California 8 algorithm has been revised and re-tested by CSST in Italy. The
         algorithm is based on a set of decision trees according to measured occupancies at both
         ends of an observed section of length 2km at intervals of at least 1 minute. Tests within the
         HERMES project in April 1995 showed that the algorithm behaved well in medium to high
         flow conditions, and was relatively insensitive to small noise signals or detector accuracy.

Issues and Limitations

AID suffers from many of the same problems that beset on-line speed control systems, regarding
the spatial resolution of the information available. However substantial efforts have now been
made to undertake data fusion from as many sources as possible to build up a complete picture of
the road situation with established AID systems. Research is currently being focused in work to be
performed as part of the 4th Framework IN-RESPONSE project, which is set to develop systems
that may use input from 9 differing detection mechanisms ranging from weigh-in-motion to
pneumatic tube counters, at six pilot sites.

A degree of standardisation in both the assessment and deployment of AID systems is slowly taking
place (CORD 1995). However their high cost implies that implementation is only likely to take
place where congestion and/or incidents regularly occur, and that therefore one of the major
directions to be taken, is in the wider dissemination of information produced by AID as opposed to
the collection of more data. Also, the existence of links from vehicles to roadside may enable a
new source of information to become available, from so called 'floating vehicles', providing spot
and link speeds (and if CAS units are available, headways and hence densities) and reducing the
need for a wide range of detectors (eg. Cremer and Putensen, 1993 and 1994).

2.3.3    Other


These systems seek to provide driver support through the provision of extra information, usually
simple hazard warnings, displayed via roadside features. Two separate candidate systems have
been considered in this section:-



The 'COMPANION' early warning system, has been developed by BMW AG through a range of
initiatives ranging from PROMETHEUS to involvement with government and local authorities.
This system consists of roadside marker posts every 50m, which are connected to a simple
information network driven by a regional control network. When an 'event' (incidents, congestion,
fog etc.) is detected by a conventional AID system, the system is activated and displays flashing
yellow or red lights, providing pre-warning for the approaching motorist (Huber et. al. 1994,
Krause 1995 and 1996). The overall cost of the system is highly dependant on local conditions (eg
road geometry, available infrastructure for power and communications systems and incident
detection facilities). A basic cost of 75kECU/km has been estimated.

Current Developments

DIATS                                       Deliverable 1

150796                                                                          DIATS/TRG/WP100

Phase 1 of the operational testing of the system is currently underway and focuses on a 9km stretch
of the A92 near Munich. Additional loops are being installed to aid existing detection systems, and
the integration of supportive detection measures such as video surveillance and image processing
are being investigated. No information is yet available on system durability or reliability, and the
effectiveness of Phase 1 is highly reliant on the efficiency of the detection algorithms used for the
system activation. In Phase 2, it is envisaged that alarm messages from equipped vehicles which are
either involved in an incident or have just passed it will provide early information. This could
shorten the reaction time considerably and thus will bring the system to its full potential and
efficiency. More detailed operational impacts of the system will be evaluated during 96-97 as part
of the DGXIII TABASCO project.

Issues and Limitations

The COMPANION system provides a valuable means of transferring information to and from
drivers and currently operating AID systems, and is set to be the mechanism through which floating
vehicle data can quickly be collected and in a suitably advanced version messages transmitted to
vehicles regarding rerouting or local speed controls. Even in its simple Phase 1 version, it provides
a valuable means of increasing the spatial frequency of information transferal to drivers, all-be-it in
the form of simple incident warnings.

The Intelligent Road Stud (IRS)


This system is marketed by Astucia Lda (Astucia, 1995), from the UK, and seeks to provide road
studs that may, like COMPANION, provide a degree of information to the driver by using the
already existing road stud system. Here, the existing studs, which reflect the light from the
headlights of oncoming vehicles and are most commonly used as lane markers, are replaced by a
new type, which contains an integral processor and coloured LED. Once activated, sensors within
the stud can detect a variety of local conditions and activate the LED appropriately according to a
preprogrammed scheme (which is unchangeable over the lifetime of the stud). In particular, the
stud LED will flash white in fog, blue in icy conditions and perhaps more importantly in these
hazardous conditions, also indicate the presence of a vehicle by displaying a flashing orange trail for
2 seconds after a vehicle passes. The stud can also warn against slow vehicles/queue formation in
all conditions, by comparing whether the speed of the passing vehicle is too slow for the road
conditions. When this occurs it will emit a flashing red light for a time dependant on the vehicle
speed (the maximum being 10 seconds for a vehicle moving at 48 kph).

The unit is self contained and powered by a rechargeable solar cell (Prynn, 1995), which has the
ability to display a hazard for several days on a single charge, and without a hazard will function for
several weeks. It is estimated that the power cell will function for approximately 100 recharges,
giving an expected lifetime of about 3 years (comparable to the lifetime of a standard unit). The
stud is produced at a price of 22 ECU/unit (compared with 5.0 ECU for a traditional unit), giving
an estimated price of 8.1kECU/km (compared with 2kECU/km). (The costs of physical
replacement of existing studs however will effectively cause the per stud price to rise above this).

Current Developments

DIATS                                       Deliverable 1

150796                                                                           DIATS/TRG/WP100

The basic unit (termed the '5 function unit') is currently under type approval by the U.K. Highways
Agency (basic structural and integrity testing), and is due for completion in Spring 1997, however,
restricted field tests are planned for the end of 1996 on the A38 (a dual carriageway). An improved
version incorporating speed detection and transmission to vehicle and roadside receivers, for
potential incorporation into an 'intelligent highway' type system, is currently at the planning stage,
and has attracted considerable interest from commercial companies.

Issues and Limitations

A suitably advanced road stud, may in essence provide the same function as the COMPANION
system, though, due to its lateral placement data is available on a per lane basis. However, in its
current form whilst it may act to warn drivers of approaching congestion without the intervention
of AID, there are several considerable behavioural concerns that require further investigation, as
drivers may become confused by which stud to 'read', or the degree of distraction that a pattern of
flashing lights on the road may cause. No work has been performed on driver reaction to the
system, or its impacts on capacity or safety.

2.3.4    High Occupancy Vehicle Lanes (HOV)


To increase the average occupancy of each vehicle (theoretically reducing total vehicle numbers),
through the provision of a separate lane dedicated for use by vehicles with over a set number of
occupants. This makes potentially congestion reducing 'ride sharing' more attractive by reducing
HOV passengers travel times.


High-occupancy vehicles (HOV) facilities have been introduced over the last 25 years, primarily in
metropolitan areas, especially in North America, to offer priority treatments to buses, van pools,
and car pools (for a review see Federal Transit Administration and Texas DoT, 1994). HOV
facilities focus on increasing the person-movement rather than vehicles-movement efficiency of a
roadway or travel corridor. Three types of facilities have been implemented to date: exclusive
facilities, separated from mixed-flow traffic by concrete barriers or physically separate lane(s):
concurrent flow lanes, placed in the peak direction of travel but not physically separated, and:
contraflow lanes, providing an exclusive lane for HOVs running in the peak direction through
removal of a lane from service in the off-peak direction, usually operating only during peak periods.
 Exclusive facilities are to a certain extent more desirable and easier to operate as the two types of
road user are kept separate, although these usually require more space and are, from efficiency and
effectiveness points of view, more costly. Appropriate measures for entry/exit from the facility are
also essential for each of the three types of facility: too few access points could inhibit usage, while
too many could interrupt the flow of HOV traffic on the facility. Ramp metering and preferential
toll treatments can be used to allow HOV users to bypass congestion points associated with mixed
flow traffic in the system.

Traditionally most facilities have a requirement of 3 or more occupants per vehicle (termed 3+),

DIATS                                        Deliverable 1

150796                                                                            DIATS/TRG/WP100

although current practice has shown that occupancy requirements should be set high and then
lowered to encourage usage. The lane could be given over to 3+ use during peak hours and 2+ at
other times, or even given over entirely to mixed traffic off peak. It is usual to allow the use of the
facility either during a 4-hour or peak-period restriction, a 24-hour restricted facility (referred to as
a fully dedicated facility) makes signing and enforcement on the facility simpler and less confusing
to HOV users but may be significantly under utilized at off peak periods.

Current Developments

i)       In Amsterdam, a reversible car pool lane was opened on the A1 in late 1993, funded by the
         national government at a cost of 3.7 Million ECU/km, over an 8Km stretch. The project
         was designed to encourage higher private car occupancy on the main road between
         Amsterdam and Almere with the lane constructed as a barrier-separated, reversible lane in
         the centre of the Highway, which acted as an HOV lane for the morning peak and a
         generally accessible lane in the evening. The project was opened for buses and car-pools
         3+ only in order to ensure free flow in the HOV lane, with resulting saving in journey time
         of 20-30 minutes. However, despite attempts to establish Park and Pool and Ride facilities,
         the degree of use remained at only 5-8% during the peak period after the first 6 months of
         operation. In late 1994, a public campaign forced the courts to rule that current legislation
         was insufficient to allow the operators to ban some highway users from facilities built with
         public money and this, combined with diminishing usage and poor public perception (HOV
         usage did not ease congestion in the morning peak, but general usage did produce some
         easing in the evening rush hour), forced the scheme to be abandoned.

ii)      The Madrid scheme (Bus-VAO, Bus Vehiculos de Alta Occupation), run by the city's
         traffic department has been fully operational since January 1995 and uses a reversible lane
         system which runs along the centre of the N-6 (a 3 lane motorway). The two-lane, 25km
         HOV, opens in one direction or the other at peak periods and allows busses and cars with
         two or more passengers direct access to the city and was constructed at a cost of ~31
         Million ECU (1.24 Million ECU/km) (Dempsey, 1996). The HOV facilities into the city
         are offered in the morning from Monday to Friday, and in the afternoons of Saturday and
         Sunday, with exit facilities offered weekday afternoons, and weekend mornings to ease
         holiday traffic. Since its opening, the HOV facility has cut rush hour journey times for a
         15km trip from 45-60 minutes to 15-20 minutes for those vehicles using the facility.
         Overall travel speed has increased from 15-20 km/h to 30-60 km/h during peak hours, with
         the facility seeing use by 8.5% of traffic during peak periods between Monday and
         Thursday and 15.1% on Fridays. The average number of passengers per car has risen from
         1.2 to 1.6. Due to the success of the system, there are plans to expand on to Madrid's two
         main ring roads (the M30 and M40) and the critical N3 Madrid-Valencia highway along
         similar lines (Diez de Ulzurrum, 1995).

DIATS                                        Deliverable 1

150796                                                                            DIATS/TRG/WP100

Issues and Limitations

One of the key issues in the definition of an HOV facility is whether the designated lane is part of
the existing road structure, or is added (typically between the carriageways). The former option
reduces the number of lanes available to non HOV traffic and potentially increasing congestion in
the mixed traffic lanes, the latter costing more and attracting additional traffic in the medium to long
term. The method of separation is also important, ie whether any physical barrier exists between
HOV and non-HOV lanes, to prevent vehicles from using the facility. If there is no barrier then
there is a drastically increased need for enforcement to avoid normal drivers moving across to the
faster HOV lane.

Compliance is essential to the operation of the facility: even if a physical barrier exists, some drivers
will still try to use it (U.S. experience shows that compliance strongly depends on levels of
enforcement), and measures have to be taken to combat this, either through the use of AVI tags for
fleet vehicles (although there is no guarantee that this will mean the tagged vehicle has the correct
occupancy), and image processing based video enforcement.

Technical considerations aside, little political willingness has been shown to back the HOV concept
in Europe and despite the success of the Madrid system, no added impetus has yet been apparent to
reconsider the approach in most of the EU nations. Part of the consideration behind this is perhaps
the very high cost associated with such a building programme and that at many sites where there
could theoretically be a benefit, there is insufficient space or adverse geometry to allow a new lane
to be added, and traffic is close to capacity so regularly that 'taking a lane' would pose an
unacceptably high risk. There is therefore no reason to believe that the next 10 years will see a
major expansion in the number of HOV facilities in the EU, with any new ones coming on
comparatively 'low risk' routes.

2.3.5    Ramp Metering


To prevent or reduce congestion on the main carriageway, by coordinating the number and timing
of merging vehicles, and in order to re-route some drivers on to other less critical junctions.


Ramp metering limits the number of vehicles entering a freeway when their insertion on the main
carriageway would lead to conflicts and added congestion. The effect is to decrease the queue on
the main carriageway and increase the (controlled) queues on the ramps. This is accomplished by
equipping the on ramp with signals signs and detectors, and fitting detectors to the main
carriageway (upstream and/or downstream the junction). The algorithm controlling the traffic
signals is based on a flow continuity equation and a desired maximum occupancy which
corresponds to an optimum of the capacity, and which has been previously calibrated. Ramp
queues must not interfere with the surrounding network, the calculated on-ramp volume must be
greater then a minimum flow, and in some cases, priority may be given for buses and, sometimes,
for car pools, through the use of a priority lane.

Traffic light control parameters may be either pre-defined from historical data (particularly in the

DIATS                                        Deliverable 1

150796                                                                          DIATS/TRG/WP100

case of recurrent congestion) or extracted from real time traffic detectors, producing a traffic
responsive system. Such systems may be broadly characterised by the presence of either co-
ordination between time steps (a global optimisation on different time steps, based on the feedback
theory, versus separate optimisations), and/or co-ordination between different ramps (for an
excellent review, see Papageorgiou, Hajsalem, Inoue et. al. and Ross, all 1990).

Current Developments

i)       The USA Demand/ Capacity strategy. Ramp metering was first developed and
         implemented in the USA in 1961, and a wide range of sites both in the US and Europe have
         since adopted the same strategy. Here, two detectors are fitted to the motorway: one
         upstream providing the volume, another downstream the occupancy, which indicates
         congestion if above a set threshold. In such a case, the volume coming from the ramp, is
         set to a minimum value or otherwise the difference between the upstream volume and the
         capacity. Similar approaches are based on the same principle, and differ either by the form
         of data used (occupancy, volume, speed), or by the number of detectors, with more
         detectors increasing the forecasting accuracy and the flexibility of the metering strategy.
         For example, a variant using the measurement of the downstream flow on the motorway
         has been tested in France in 1978 by INRETS on the Boulevard périphérique de Paris
         (Morin, 1981).

ii)      A traffic responsive strategy on isolated intersections has been developed and tested by
         Wootton Jeffreys Consultants in the UK (Owens and Schofield, 1988). This strategy was
         first implemented in 1986, on six entry ramps on the M6, located in the West Midlands
         Conurbation, and has since been tested on the Boulevard Peripherique and on the
         Coentunnel in The Netherlands. The peculiarity of this method is in its use of speeds,
         recorded by a downstream detector, the monitoring of which allows a good estimation of
         the capacity for the current traffic conditions. This approach does however require an
         accurate validation of the speed-density relationship.

iii)     The Rijkwaterstaat Strategy (the Netherlands). This approach has now been fully
         implemented and tested on the Coentunnel in Amsterdam and on Delft-Zuid
         (CHRISTIANE, 1990). An important objective in this application is the diversion of
         vehicles away from a number of sensitive ramps. The strategy is "one car by green",
         leading to a short green time (5-15 seconds), which allows the passage of only one vehicle
         per lane, with the cycle time computed in order to adjust the rate of passage to the required
         rate (the capacity minus the (smoothed) upstream motorway demand) every minute.
         Results so far (Middleham, 1994, EUROCOR, 1995, and Taale and Van Velem, 1996)
         have indicated that the numbers of drivers merging the "coarse way" has decreased from
         15 to 4%, with only a few percent violating the signal scheme. No notable change has been
         observed regarding accident rates, however, the frequency of flow breakdowns and their
         spatial extent has been reduced, giving rise to a flow increase on the mainline of 2-3%, with
         a resultant reduction in travel time. It is also notable that 33% of vehicles have diverted to
         use neighbouring less critical ramps. This strategy is also due to be implemented during
         1997/8 on two test sites in Germany: in Berlin on the urban motorway A100, and in the
         Ruhr area (Northrhine-Westphalia) on the A40 motorway.

iv)      ALINEA and METALINE. Two French ramp metering strategy (respectively on isolated

DIATS                                        Deliverable 1

150796                                                                         DIATS/TRG/WP100

         intersections and on several coordinated ramps) have been developed by INRETS, based
         on both upstream and downstream conditions. The pilots of these systems were installed in
         1985 (for ALINEA) and in 1989 (for METALINE), and are now fully operational with
         different tests and evaluations having been performed on the Boulevard Peripherique de
         Paris and on the ringway at Amsterdam. The on-ramp volume is controlled in order to
         produce a desired occupancy downstream.
         *       For ALINEA, a 16% saving in journey time has been noted for peak hour traffic,
                 with the duration of congestion being caused by the morning peak reducing by 50%
                 (Hajsalem et. al., 1988, and Papageorgiou et. al., 1989). The cost of the system,
                 traffic detectors, modification of the traffic lights in 1985 was rated at
                 approximately 8 kECU/ramp.
         *       For METALINE, a link between the different ramps is established in order to
                 develop an optimal mix of ramp metering timing to ensure a desired travel speed
                 and times on the mainline. The algorithm is based on Linear Quadratic Control (the
                 criteria is quadratic, the dynamic model is linear). So far savings of 20 % have been
                 observed in travel time, and 13% in the duration of flow breakdowns
                 (CHRISTIANE, 1991). Due to the degree of coordination required between
                 ramps, installation cost is somewhat higher at 32 kECU/ramp. Non-linear control
                 has also been developed and successfully tested in simulation, but has not been yet
                 implemented, the algorithm being rather slow to run (Bhouri, 1990).

Issues and Limitations

The use of ramp metering in the EU is currently growing either as an isolated function, or included
in an integrated motorway management package. In this latter respect, research is currently being
focused in the 4th Framework D'ACCORD project, which is set to conduct tests on both the
peripherique (35km and 70 on and off ramps) and the motorway network around Paris (highways
A1 to A13) over a total of 515 Km. Initial off-line tests will be conducted using the METEOR
model from the EUROCOR project, and in the on-line phase, new techniques developed for
detecting and extrapolating for missing data and traffic forecasting and travel time estimation, with
particular test sites identified as the Peripherique and the A6 west, and motorway to motorway
control on the entrance to the A6-Peripherique.

Although there is little commonality in terms of goals or approaches among the sites where ramp
metering is implemented, one particular use of all of them may be in its interaction with cooperative
driving systems where the merging sequence could be controlled as a function of the lane density of
oncoming equipped vehicles (which may pass the junction grouped as a loose platoon).

DIATS                                       Deliverable 1

150796                                                                          DIATS/TRG/WP100


It is instructive to view how similar approaches to those presented in Section 2 have fared in
differing financial and political environments that may be found outside of Europe, and how these
systems may compete with those produced within the EU. Of particular relevance here are the
body of projects represented by VERTIS in Japan, and ITS America in the U.S.A (see for example,
U.S. DoT and FHWA, 1996).

3.1      U.S.A.

The most ambitious project of relevance to this review is that of the National Automated Highway
System Consortium program (NAHSC) (Saxton and Bishop, 1994), part sponsored by the
Department of Transportation, aimed at the construction and test of a range of AHS prototypes as
summarised in Section 2.1. This project is based on the findings of earlier studies which concluded
in 1995, dedicated to investigating three main areas relevant to AHS, collision avoidance, human
factors, and 15 sub studies dedicated to evaluating issues, risks, and concerns related to
deployment. These 15 studies, termed the Precursor Systems Analyses (PSA) proposed several
paths to the implementation of a fully cooperative AHS architecture that would conclude with the
formation of road train type system, with full inter vehicle and vehicle to roadside communications.
 Many of these paths incorporated a step wise introduction of the individual operational elements
such as those reviewed in this deliverable, and most notably rejected the concept of mixed
equipped-non equipped traffic lanes (Bishop, 1994 and Harding et. al., 1995).

The next phase to be undertaken by this consortium is that of systems definition, which was started
in late 1994, in order to select and document the approach to be evaluated in the third and last
phase. This is similar to the PROMETHEUS programme in that it brings together government
with industry and academia, to develop and test operational prototypes suggested by the PSA.
This phase is scheduled for completion by 2002, with intermediary stages including presentation of
the test vehicles at a major open day in 1997 (Quinlan and Gouse, 1995), subsequent system
evaluation and selection, and the construction of a second phase, advanced prototype. The final
operational test and evaluation phase, is to start before 2004.

The technical background for these developments are based on earlier (and ongoing) research work
conducted by PATH which over the last 5 years has established a wide range of fully cooperative
prototypes, as well as highway management and communication software (Shladover, 1995).
Initial work by this consortium (Chang et. al., 1991) has successfully shown that it is possible to
build a suitable control system that can maintain an inter-vehicle spacing of 10m with little variation
(+/- 5% at speeds up to 25 m/s), provided that the deceleration exhibited by the lead vehicle stays
within acceptable bounds (> -1 m/s ). Subsequent experiments (Chang et. al., 1994) using similar
acceleration limits have found that the following controller can be extended to 35 m/s, with little
increase in the spacing variation. Perhaps a more important result however, is the extension of the
controller system to cope with an extended platoon (4 vehicles) where it was found that the last
vehicle in the convoy could be sufficiently well controlled to restrict its headway fluctuations to
under 20%. Lateral control using image processing and 'magnetic nails' has also been tested with a
maximum lateral deviation at 60 km/h measured at 0.15m. Lastly, and most significantly have been
tests on platoon joining manoeuvres with the lead vehicle travelling at 100 km/h, joined by the
second vehicle from an initial distance of 30m to a final distance of 4m. Throughout the manoeuvre
the maximum speed profile tracking error was less than 0.5 km/h. Although all these technical

DIATS                                       Deliverable 1

150796                                                                          DIATS/TRG/WP100

developments are a long way from the final system design, progress made over the last 5 years
would seem to indicate that the final architecture is feasible in the developmental time frame

In parallel to these developments the National Highway Traffic Safety Administration (NHTSA)
has its own separate research program in cooperative driving, dedicated not only to development
and testing but also to the establishment of performance standards (Leasure et. al., 1994). This has
included work into an ADR (portable Driver Performance Data Acquisition System for Crash
Avoidance Research - DASCAR), driver status monitoring, MMI and vision enhancement (a major
state of the art review conducted in 1994), as well as advanced prototypes for the standard range of
CAS functions such as Lane detection (being pursued by Rockwell), ICC systems (UMTRI and
Leica, under the $1.8 m FOCAS project), with supplementary work regarding user acceptance
being performed by the Ford Motor Company. Within FOCAS, 10 Chrysler cars have been
equipped with ranging sensors and incorporate GPS location equipment in order to investigate how
the functions provided by an AICC can be extended to produce a rear-end collision avoidance
system. Variables under investigation include driver age and driver behaviour before and after
installation, as well as possible problem areas such as sensor performance during merging and
weaving manoeuvres, and simulation based studies in to traffic impacts and emission reduction.

Of the major vehicle manufacturers and suppliers, Delco electronics (part owned by General
Motors) have been most notable in prototyping the FOREWARN system (Schumacher et. al.
1995), based on the radar sensors to both the front and rear which provides the driver with a
collision warning, with a good deal of research appearing to have been performed into the
appropriate media and message timing. The system is preprogrammed with typical data regarding
relative braking distances and rates for 'average' drivers (which can be set between predefined limits
by the driver) and if these thresholds are exceeded a warning is given both audibly and visually
through a head up display. Once the driver decelerates, these warnings are extinguished, but if the
driver does not act and the vehicle continues to close to an 'unsafe' distance then a second stage is
activated with a verbal message, a flashing symbol, and feedback through the accelerator pedal.
The system is also supplemented by the use of a 'tailgating' symbol to prevent the driver from
getting too close. The distance and relative speed thresholds used in defining these warnings may
also be extended according to information from onboard sensors regarding weather (windscreen
wipers), tyre pressure, driver attention (radio on/off) in order to allow a safety margin for error. As
yet commercial exploitation has not proceeded and it is believed that this is due to commercial
strategy considerations that include consumer price resistance, competitive development, as well as
potential unresolved legal issues, and user acceptance.

Among developments of infrastructure based systems such as those reviewed earlier, the most
extensive is an on-line speed system called TRAVEL-AID, scheduled to open in 1996 over a 65
km length of Interstate Highway 90, in Washington State, and believed to be the first of its kind in
the U.S.A.. It will incorporate a number of infrastructure and in-vehicle methods of informing
drivers of adverse weather and traffic conditions with information being displayed by the use of
VMS and transmitted to in-vehicle RDS/TMC units. The estimated cost of the project is expected
to be approximately 1.6 Million ECU funded through NHTSA.

With the exception of on-line speed control, virtually all the other main infrastructure based and
related system highlighted in this deliverable have seen just as widespread development and testing
in the U.S.A. as the EU. Clearly there are many dozens of systems, all generally capable of the

DIATS                                       Deliverable 1

150796                                                                         DIATS/TRG/WP100

same degree of performance as their EU counterparts. The most well known of these are perhaps:

i)       The AID 'Minnesota' algorithm (Stephanedes, 1991, 1993), which focuses on the
         elimination of false alarms by preprocessing the data received in order to attenuate the
         noise. This has been tested on a 9 km segment of the I-35W in Minneapolis over a 6-
         months period and has produced a mean time to detection of 3 minutes and a 0.2% false
         alarm rate with an 80% detection rate.

ii)      In ramp metering, the "Percent-occupancy Strategy" is noteworthy as being the only system
         that needs only one upstream detector and uses a dynamic equation to define and compute
         the optimum on several time steps.

iii)     The most notable difference with the EU lies in the degree of use of HOV systems which
         are now widespread in numerous metropolitan. As of autumn 1992, there were some 49
         HOV facilities in operation on either freeways or in separate rights-of-way in 22 North
         American metropolitan areas and recently, several large scale overview reports have been
         published, focusing on synthesising the impact analyses conducted on major installations, to
         assess the effectiveness and flaws in the HOV approach to congestion management (eg.
         Federal Transit Authority and Texas DoT, and Virginia Transportation Research Council,
         both 1994).

3.2      JAPAN

Within Japan, it is somewhat easier to keep track of the major development programmes and
sources of funding, however, in all but the 'high profile' areas such as AHS, it is perhaps more
difficult to establish a consensus as to how widespread developments have been and operational
details. This is perhaps understandable in view of the obvious barriers presented by both language
and geographical location. Most ITS developments in Japan in recent years have been undertaken
under the coordination of VERTIS (Komoda, 1995). This organisation has been responsible for
the establishment during 1995 of the long term research plan for Japanese involvement in ITS, and
has established strong ties with differing branches of central government, each of which is generally
responsible for the coordination and promotion of developments on a particular theme. Within
these themes, several work areas, often with some overlap between them, are apparent. The main
features are described below.

Work on ATIS (Advanced Travellers Information Systems) and ATMS (Advanced Traffic
Management Systems) is being focused in the ARTS project (Advanced Road Transportation
System), which started in 1989 with the backing of the Ministry of Construction. This project is
concerned with establishing vehicle-roadside communications links and suitable systems for both
lateral and longitudinal control and warning by the year 2000, with full integration into a
communications infrastructure (itself starting tests in 1996). This is intended to produce an
Automated Highway System (AHS) by 2005. Of particular note are developments in:-

i)       Upgrading current AID systems by establishing a direct link with equipped vehicles and
         using them as floating probes (similar to the methodology being explored in the
         SOCRATES line of projects in Europe and ADVANCE in the U.S.A.), called RIDIS

DIATS                                       Deliverable 1

150796                                                                         DIATS/TRG/WP100

         (Road Incident Detection and Information Service).

ii)      Rapid developments being made in establishing a fully automated lateral guidance system
         using image processing and magnetic nails embedded in the road, where a prototype system
         has been tested with a target 1 second headway (Nakamura and Yoshikai, 1995). All the
         major motor manufacturers are represented, with road-to-vehicle communication
         infrastructure established based on Leakage Coaxial Cable (LCX) technology. Under tests
         six sections of LCX each 500m long were used, receiving information from vehicles
         regarding their positions and speeds and sending speed instructions and hazard warnings
         (road bending etc). Demonstrations undertaken so far have comprised a run starting at 40
         km/h, accelerating to 60 km/h, then down to 50 km/h and finally to a stop with a constant
         5m spacing.

iii)     Guidelight (Imacho et. al., 1995) is similar to BMW's COMPANION system, however in
         this case the roadside beacons have been fitted with a small radar detector and the system is
         primarily intended for deployment on rural roads with low visibility, where the systems
         warning lights will activate when a vehicle nears a curve, to alert the driver of potential
         danger. The system has been under test since 1994 and current developments are being
         made to incorporate the communication technology and protocols being established in
         RIDIS, to develop essentially a system equivalent to a 'Phase 2' COMPANION.

Meanwhile, research into AVCS (Advanced Vehicle Control Systems) starting in 1991 has mostly
been promoted through the SSVS (Super Smart Vehicle System) and ASV (Advanced Safety
Vehicle) projects backed by the Ministry of International Trade and Industry. The SSVS project
has established a target of 2020 for the introduction of their AHS system, which is based on a long
history of developments in this area reaching back to inductive wires as a basis for lateral guidance
in the 1960's, and roadside beacons in the 70's (Tsugawa, 1995).

Work in the ASV project (started in 1991) is projected to end in 1996 and has focused on 20 work
areas, prototypes of which were all produced in 1993, and are set to be demonstrated to the public
in 1996. Relevant areas here include all the typical CAS functions found in Europe and the U.S.A.,
for example:-

*        Mitsubishi (Watanabe et. al., 1995) and Mazda (Butsuen et. al., 1994), have both
         developed and tested an ICC with a set 2 second headway, based on the use of a laser
         radar, supplemented by image processing of the road scene to determine the relevance (and
         lane) of each of the detected vehicles. Additionally, this unit will also provide the driver
         with an early warning, if the vehicle needs to be decelerated more than the amount
         accessible by using the accelerator pedal. The Mazda unit has only just started to undergo
         restricted road tests, however the Mitsubishi unit is now being offered on a commercial
         basis on the Japanese market version of the Diamante.

DIATS                                       Deliverable 1

150796                                                                           DIATS/TRG/WP100

*        Nissan, have adopted a differing approach to CAS design (Satoh et al., 1994), in that they
         have opted to use a 3 beam fixed laser unit (similar to Leica's ODIN series sensor), and
         have based the deceleration profiles adopted by the unit (ie. when to decelerate, by how
         much etc.) on data collected on real driving behaviour using an instrumented vehicle. This
         has the added advantage of matching the driver's expectation of driving performance, but
         raises some legal questions regarding the absolute safety of the device.

Additional areas being pursued within ASV have included: driver status monitoring, based primarily
on monitoring of the steering wheel angle: all round proximity warning in darkness: a rear view
collision warning system developed by Nissan, that sounds a warning if a vehicle is approaching too
fast from the rear and flashes rear view warning lights to alert the approaching driver (Tomioka et.
al., 1995): lane change and lane deviation warnings: and ADRs (Miyazaki, 1995).

Relevant work has also been performed under the VICS project (Vehicle Information and
Communication System), where a country wide system of roadside radio communication beacons
has been developed and tested. This system is about to undergoing the first phase of deployment
(due for completion in 2002), being fitted every 2-4 km along expressways in 8 prefectures
(districts) covering 9% of the country and reaching a projected 35% of the vehicle ownership. The
projected second phase (from 2002 to 2008) will increase the coverage to 36% of the country and
48% of the ownership (Mizoguchi, 1995).

It is understood that several other developments of note are being pursued outside of the VERTIS
development framework:

*        The Intelligent Delineator System (Kajiya et. al., 1995) tested in early 1995, which is similar
         to the IRS, but with warning lights being placed on top of a small pole, that is sited at the
         lane edge. Here the functionality is broadly similar and warning lights may be activated by
         bad weather or an integral radar unit configured to detect stationary vehicles.

*        A video based AID system (seemingly the first of its kind in Japan) similar to INVAID has
         been installed and tested in late 1994 primarily with the aim of detecting stationary vehicles
         on curves and in tunnels, where it has achieved a 96% detection rate (Taniguchi et. al.,

DIATS                                        Deliverable 1

150796                                                                         DIATS/TRG/WP100



Identification of specific users and providers is not essential at this stage of the project, and is
required mainly in order to identify the varying levels of interest that both of these groups would
have in system implementation. This provides, in parallel to work performed in WP900 (Liaison
with, and participation of, owners and user groups), one of the starting points for activity 150
which deals with the likelihood of deployment, incorporating cost and manufacturers' viability
criteria that have been outlined where possible in section 2.

A summary of the major participant categories examples and major issues associated with them is
as follows:

i)       Public agencies. Governmental agencies include primarily those of the EU, national
         governments, and county-level and local jurisdictions.
         a)     EU Directorates (e.g. DGVII): Pan-national development, standardisation,
                resource conservation, global competitiveness policy directives regarding research,
                co-ordination of research, fostering research/industry co-operation.
         b)     National government interests: Co-ordination of national highway construction and
                operational strategies (e.g. the Highway Agency in the U.K.), industrial
                development, competitiveness national standards (e.g. Departments of
                Trade/Industry), resource conservation, policy formulation and response to user
                and public needs, legal and regulatory agencies with nation-wide responsibilities
                (e.g. Departments of Transport).
         c)     County/department/land, and local governmental agencies: Transport and land-use
                agencies with concerns which are a component of national systems and which must
                be coordinated with initiatives at national levels. These groups will have interests
                particularly at the interface of interurban motorways and local roads and related
                transport facilities. Also, where installation and operation of ATT systems occur
                within these jurisdictions, concerns over environmental impacts such as air pollution
                and visual effects on local areas may have to be considered.

ii)      Trade and Industry Representatives. These organisations represent the interests of groups
         of companies. Outlined below are those representing motor vehicles, freight, infrastructure
         and electronic manufacturing, and operating companies.
         a)      Motor vehicle manufacturers: Adaptation of design and possible retrofitting of
                 existing fleet to accommodate ATT equipment, cost, marketing and public
                 acceptance issues.
         b)      Freight and Public Transport operators, such as the Freight Transport Association:
                  Installation of ATT equipment and driver training and compliance issues,
                 cost/productivity, driver performance and legal/regulatory issues.
         c)      Infrastructure construction companies, and electronic equipment manufacturers and
                 operators: Design, construction and maintenance issues, commercial opportunities,
                 industry-wide standardisation, investment in research and development, legal and
                 regulatory issues.

iii)     End-user representatives.    This group may be typified by the various automobile

DIATS                                       Deliverable 1

150796                                                                            DIATS/TRG/WP100

         associations, and perhaps, operators' associations, although some overlap occurs with those
         mentioned above. The group is expected to comprise groups that may be important in
         representing the interests of end-users of the ATT systems. End-users are considered here
         to be primarily drivers and vehicle occupants. Some key issues could include the financial
         ability of certain drivers to afford the ATT systems, the ability to effectively use the
         equipment (possibly elderly drivers), and requirements for commercial driver training.

iv)      Legal and regulatory. It is assumed here that these groups would be represented within the
         governmental agencies mentioned above. However, it may be appropriate to include a
         separate, independent legal perspective in response to the different legal systems
         throughout Europe and requirements for standardisation.

Currently, contacts are being established with organisations formed from of all the major interest
groups, and a framework formulated for requirements and criteria to assist selection of the most
appropriate representatives. As the project proceeds the list of participants will evolve from
consideration of demonstrated interest, input contribution and potential for maximum benefit to the
manufacturer and user populations.

A detailed interpretation of how the concerns of each of these groups are addressed by the systems
in this deliverable is the focus of activity 150, and as such will not be addressed in detail here. It is
possible to summarise the nett potential benefits that each group could receive however, (building
on earlier examinations by Morello et. al., 1994 and CORD, 1994a), and these are displayed in
Table 1. This shows the basic breakdown of the systems: cooperative: in vehicle devices:
infrastructure based and 'other', and also shows how for example for cooperative systems,
government agencies may find that the benefits from their investments may prove negative if the
system fails because of lack of acceptance on the part of the users. Also regulatory concerns may
be negative if additional legislation is required.


As has been stated in the introduction to this deliverable, a vital element in considering the test
systems within this project is also to establish a measure of foresight as to how these may change.
These changes may take place both technically, and perhaps more importantly, from the stand point
of driver behaviour (and usage), may change from that which we currently observe and base our
work on. The technical changes likely to occur are to be examined in the next two activities within
this workpackage and will be summarised in Deliverable 2, and behavioural elements investigated in
WP600. It is instructive however even at this early stage to summarise some of behavioural issues
that have been highlighted in the earlier sections, in order to establish a framework for additional
data requirement that may be obtained from other research activities. The primary issues are as

i)       For in-vehicle systems although user preferences have now been established for MMI,
         'short term' user reaction, in terms of changes in driving headway for example, seem to be
         dependant on the method of assessment (real road tests vs simulator based). Before any
         degree of reliance can be attached to these outputs the interplay of observed behaviour and
         methodology must be clarified, with priority being given, where available, to results
         produced from real world tests.

DIATS                                        Deliverable 1

150796                                                                          DIATS/TRG/WP100

ii)      It is unclear how permanent 'short term' changes actually are, in that current tests have
         exposed the driver to the system under test for little more than 10 hours or so, and even
         then only under a restricted set of conditions regarding flow, weather and geometrical
         conditions of the roads used. Available data on behavioural modifications therefore must
         be viewed as having little transferability from the scenario under which it was collected.

iii)     It is as yet uncertain, what effect codriver systems may have on the overall degree of driver
         attentiveness and reaction time, and how far the driver may be lulled in to a 'false sense of

iv)      It is possible that a degree of risk homeostasis will occur, with the safety benefits a system
         produces in one behavioural indicator (eg. headways or speed differentials), being offset by
         a change in another (eg. a greater propensity to change lane, higher occupancy in 'faster'

v)       The degree of belief a driver has in any cooperative system will govern his degree of
         compliance with any suggested course of action. This in-turn may effect the success of the
         system, in-turn effecting belief and compliance. For example a driver will not comply with
         a seemingly inappropriate variable speed limit (unless forced), and this in turn may make the
         posted limit more inappropriate.

Increasingly, attention is turning to these problems, with issues such as (iii), undergoing
investigation in the UDC project, and new data set to become available on measures relevant to
issues (iv) and (v) from increasing use of instrumented vehicles and more advanced data fusion
techniques used as part of the development of AID and Ramp metering tests (eg. Yang et. al. and
Yang and Yagar, 1994).

While it is possible to obtain some data on dynamic behavioural changes such as those presented
above, one must also consider how a drivers behaviour will be influenced by systems that have
earlier been installed at any test site, where it is likely that a complex inter-relationship will form
between the penetration and effectiveness of in-vehicle systems and the success of the more wide
spread infrastructure based systems. For example, with the continuing growth and success of
infrastructure systems, drivers may see little point in purchasing in-vehicle systems which may only
aid personal comfort at low to medium flows. This factor will result in a very slow implementation
to the market, and although eventually AICC equipped vehicles may reach a sufficient penetration
that the accrued benefits increasingly outweigh those available from infrastructure units, this benefit
could occur far sooner if the in-vehicle equipment is marketed forcefully, not only at a low price,
but also as part of a complete package of in-vehicle measures.

Clearly, these issues are of vital importance to our understanding of the deployment scenarios
which this project seeks to define, and is intricately linked to not only the marketing policy of
manufacturers but also to the long term goals of governments and infrastructure owners. High
emphasis is currently being placed therefore in establishing a close working relationship between
these bodies and the DIATS project through activities in WP900, in order to clarify possible
'background' conditions that will be required in the definition of the project technical work
programme in WP300.

DIATS                                        Deliverable 1

150796                                                                      DIATS/TRG/WP100

 System                                                                System        Participants
                                    Public Agencies     Manufacturer   Transport     Traffic Control   Individual   Regulatory
                                                                       Operators     Operators

 Cooperative Driving                ++/--               ++             ++            +                 ++           --

 AICC                               +                   +              +             +                 +            -
 Vision Enhancement                 0                   0              +             +                 ++           0
 Driver Status Monitoring           +                   +              ++            ++                +            0
 and Support
 Dialogue Management                +                   +              +             +                 +            0

 On-line Speed Control              +                   0              0             +                 +            0
 AID                                +                   0              0             +                 0            0
 Other                              +                   +              0             +                 0            0

 HOV                                -                   +              +             0                 -            -
 Ramp Metering                      0                   0              0             +                 0            0

+         indicates a perceived benefit
0         indicates little perceived benefit
-         indicates a perceived dis-benefit


DIATS                                          Deliverable 1

150796                                                                          DIATS/TRG/WP100


This deliverable has presented a range of cooperative ATT measures that may be used to increase
driver safety and comfort, while minimising travel times and reducing both the likelihood and
magnitude of congestion on interurban roads. The systems considered have been a mix of those
that are (or may soon be) available for implementation either in-vehicle or off-vehicle, as part of the
infrastructure of an interurban traffic management system.

The review has summarised the current state of the art of these systems and presented a list of
benefits and limitations for each, while additionally attempting to predict the degree to which each
may have progressed in the near future (an issue to be further evaluated in Deliverable 2).

The review has found that infrastructure based systems have now generally reached maturity and
that focus is shifting from their development to their widespread deployment. In-vehicle systems
on the other hand, although having been developed to an advanced prototype stage have yet to be
deployed. The components of a true cooperative system (though stopping short of the AHS
concept) have now started to be developed in a range of projects, and may eventually see vehicles
transmitting data to roadside collection points for use in infrastructural decision making processes.

An examination of international developments has shown that the EU has stressed both the
development of individual infrastructure and in-vehicle systems more than Japan or the U.S.A. The
Japanese however have generally focused more on establishing links between the two areas in the
deployment of a road side to vehicle communication system, that may be used both in an AHS and
in improving the information available to existing infrastructure measures. Alternatively, in the
U.S.A far greater emphasis has been placed on establishing the necessary elements for an AHS,
with the 'low tech' options of AICC for example having, until recently, seen less emphasis.

Lastly additional data sources for use in this project have also been identified, both technically and
perhaps more importantly, for behavioural shifts that may be induced by systems planned for
widespread deployment in the next few years. Additionally a number parallel investigations have
been identified outside of the EU that have a direct overlap with the goals and objectives of this
project, and contacts established in order to benefit from their experiences.

DIATS                                       Deliverable 1

150796                                                                      DIATS/TRG/WP100


Astucia Lda. (1995). Intelligent Road Studs: The Future in Road Safety. Promotional Material,
Astucia Lda. Doncaster, UK.

Augello, D. J. (1993). Description of Three PROMETHEUS Demonstrators Having Potential
Safety Effects. Proc. of the Thirteenth International Technical Conference on Experimental Safety
Vehicles. IRRD 864606.

Bangemann, M. et. al. (1994). Europe and the Global Superinformation Highway. Report to the
European Council. Commission of the European Communities, Brussels, Belgium.

Becker, S., Bork, M., Doriβen, H. T., Geduld, G. and Hoffman, O. et. al. (1994). Summary of
Experience with Autonomous Intelligent Cruise Control (AICC). Parts 1 and 2. Proc. of the 1st
World Congress on Applications of Transport Telematics and IVHS, pp 1828-43. ERTICO.

Becker, S., Brockmann, M., Mertens, A., Niu, R. and Sonntag, A. (1995). User Acceptance and
Willingness to Pay for Advanced Driver Support Systems. TRAFFIC Technology Europe '95.

Benz, T. (1994). Checking ICC in a Realistic Environment. Proc. of the 1st World Congress on
Applications of Transport Telematics and IVHS, pp 925-32. ERTICO.

Bhouri, N., Papageorgiou, M., Blosseville, J.M. (1990). Optimal control of traffic flow on
periurban ringways with application to Boulevard Périphérique in Paris. Proc. 11th IFAC World
Congress, Dusseldorf.

Bishop, R. (1994). Results of Precursor Systems Analyses of Automated Highway Systems. Proc.
of the 1st World Congress on Applications of Transport Telematics and IVHS, pp 1959-66.

Brookhuis, K. A., Schrievers, G., Tarriere, C., Petit, C. and Chaput, D. (1991). Monitoring Driver
Status Through In-Vehicle Parameters. Advanced Telematic in Road Transport. Proc. of the
DRIVE Conference, Brussels. Vol 2, pp 1516-25. ELSEVIER.

Brookhuis, K. A. (1993). Detection, Tutoring and Enforcement of Traffic Rules Violations-The
DETER Project. Proc. of the IEEE-IEE Vehicle Navigation and Information Systems Conf. pp
698-701. IEEE.

Broqua, F. (1992). Impact of Automatic and Semi-Automatic Vehicle Longitudinal Control on
Motorway Traffic. Proc. of the Intelligent Vehicles Symposium, Detroit, IEEE.

Busch, F. (1990). Incident Detection. Concise Encyclopedia of Traffic & Transportation Systems,
(Editor: Papageorgiou, M.), Pergamon press, 1990, pp 219-25.

DIATS                                     Deliverable 1

150796                                                                        DIATS/TRG/WP100

Busch, F., Cremer, M., Ghio, A. and Henninger, T. (1994). A Multimodel Approach with Fuzzy
Reasoning for Traffic State Estimation and Incident Detection on Motorways. Proc. of the 1st
World Congress on Applications of Transport Telematics and IVHS, pp 1245-52, ERTICO.

Butsuen, T., Doi, A. and Sasaki, H. (1994). Development of a Collision Avoidance System with
Automatic Brake Control. Proc. of the 1st World Congress on Applications of Transport
Telematics and IVHS, pp 2079-86. ERTICO.

Chang, K.S., Li, W., Devlin, P., and Shaikhbahai A. et. al. (1991). Experimentation with a Vehicle
Platoon Control System. Proc. of the VNIS Conf., pp 1117-24.

Chang, K.S., Hendrick, J.K., Zhang, W.B., Varaiya, P. et. al. (1994). Automated Highway System
Experiments in the PATH Program. IVHS Journal, 1, pp 63-87.

CHRISTIANE (1990). Isolated Ramp metering. Deliverable 5 - Appendix 1A. DGXIII, Brussels,

CHRISTIANE. (1991). Isolated Ramp metering Strategies: Real-life study in France. Deliverables
7A, 7B and 13. Ed. Hajsalem, H. et al. DGXIII, Brussels, Belgium.

Cohen, S. (1993). Surveillance du trafic autoroutier - tests en Ile-de-France avec le logiciel DAISI.
Recherche Transports Sécurité, no 37, pp 15-22.

CORD (1994). Recommended Definitions of Transport Telematics Function and Subfunctions.
Deliverable D004 - Part 3. Edited by Ryd, P. O. DGXIII, Brussels, Belgium.

CORD (1994a). Guidelines for the Assessment of Transport Telematics Applications in Driver
Assistance and Cooperative Driving. Deliverable AC07 - Vol. 2. Edited by Morello, E. DGXIII,
Brussels, Belgium.

CORD (1995). AID Task Force: Guidelines for Implementation of Automatic Incident Detection
Systems. Deliverable AC10 - Part 2. Edited by Zhang, X., Busch, F., and Blosseville, J.M..
DGXIII, Brussels, Belgium.

CORD (1996). Driver Assistance and Co-operative Driving: Results of the ATT Cross-Project
Collaborative Study. Deliverable AC16 - Vol. 2. Morello, E. et. al. DGXIII, Brussels, Belgium.

Cremer, M. (1981). Incident Detection on Freeways by Filtering Techniques. Proc. of the 8th IFAC
World Congress, Kyoto, Japan, Pergamon Press.

Cremer, M. and Putensen, K. (1993). Monitoring Traffic Density Profiles Using Measurements
from Loop Detectors and Vehicle Trip Data. Conf. on Advanced Traffic Management Systems, St
Petersburg, Florida.

Cremer, M. and Putensen, K. (1994). An Approach for Traffic State Estimation Using Invehicle
Information. Proc. of the 7th IFAC/IFORS/IFIP Symposium on Transp. Systems, Tianjin, China.

Cremer, M., Morello, E., Sala, G. and Schrieber, S. (1994). Experimentation of AID Algorithms in

DIATS                                      Deliverable 1

150796                                                                     DIATS/TRG/WP100

a Real Environment. Proc. 1st World Congress on Applications of Transport Telematics and
IVHS, pp 1143-50, ERTICO.

Danish Ministry of Transport (1995). TRIM - A part of the Intelligent road. Road Directorate,
Danish Ministry of Transport.

Daviet, P. and Parent, M. (1995). Platooning for small public Urban Vehicles. 4th Int. Symp. on
Experimental Robotics.

Dempsey, P. (1996). Managing Madrid's Traffic Mania. Traffic Technology International,
Feb./Mar. pp 28-32.

DGVII. (1994). Transport RTD Work Programme - Additional Information Document: Road
Transport. p41. DGVII, Brussels.

Diez de Ulzurrum, J. (1995). The Bus Vao System. Proc of the 2nd. Proc. of the 2nd World
Congress on Intelligent Transport Systems, pp 408-12. VERTIS.

EUROCOR. (1995). Deliverable 11A. Experimental Results and Comparative Analysis. Ed.
Middelham, F. et. al. DGXIII. Brussels, Belgium.

Evans, R. (1995). Head Up Displays for Vision Enhancement. Presented at the AUTOTECH95
Conference, Paper C498/8/197. IMechE, London, U.K.

Fairclough, S., Planque, S., Martinez, D. and Brookhuis, K. (1994). Behavioural Responses to a
Driver Impairment Monitoring System. Proc. of the 1st World Congress on Applications of
Transport Telematics and IVHS, pp 2033-8. ERTICO.

Federal Transit Administration and Texas Department of Transportation (1994). High Occupancy
Vehicle Project Case Studies - Historical Trends and Project Experiences. U.S. DOT.

Fincham, W. and Fowkes, M. (1994). SAMOVAR Uses a Specialised Vehicle Data Recorder to
aid Traffic Accident Reconstruction. Proc. of the 1st World Congress on Applications of Transport
Telematics and IVHS, pp 1975-82. ERTICO.

Fincham, W. (1996). Watching How You Drive. IEE Review, May 1996. pp 97-100.

Fisher, G., Morin, J. M. and Schofield, M. (1994). Area 4 Report: Integrated Inter-Urban Traffic
Management. DGXIII, Brussels.

Groeger, J. A. and Kuiken, M. J. (1994). From PSALM to HIM: Developments in Instructional
Support. Proc. of the 1st World Congress on Applications of Transport Telematics and IVHS, pp
2024-31. ERTICO.

Hajsalem H., Blosseville, J.M., Davée, M.M. and Papageorgiou M. (1988). ALINEA; un outil de
régulation d'accès isolé sur autoroute - Etude comparative sur site réel. INRETS Report n° 80,
INRETS, Arcueil, France

DIATS                                     Deliverable 1

150796                                                                      DIATS/TRG/WP100

Hajsalem, H. et. al. (1990). On-Ramp Control - Local. Concise Encyclopaedia of Traffic &
Transportation Systems (Editor: Papageorgiou, M.), Pergamon press, 1990. pp 295-300.

Harding, J., Lay, R., McHale, G. and Stevens, W. (1995). Summary and Assessment of Findings
from the Precursor Analyes of Automated Highway Systems. The MITRE Corporation.

Hardman, E. (1996). Motorway Speed Control Strategies using SISTM. Proc. of the 8th Int. Conf.
on Road Traffic Monitoring and Control. pp 169-72. Conference Publication No. 442, IEE.

Helmuller, R. (1995). Pkw halt automatisch sicheren Abstand zum Vordermann. VDI-Nachrichten
Nr. 29.

Hoogen, E. and Smulders S. (1994). Control by variable speed signs: Results of the Dutch
experiment. Road Traffic Monitoring and Control. Conference Publication No. 391, IEE.

Huber, W., Krause, G. and Keller, H. (1994). Technical Concept and Field Trial Planning of the
Roadside Warning and Information System COMPANION. Proc. of the 3rd Int. Conf. on Safety
and Environment in the 21st Century, Tel Aviv, Israel.

ICARUS. (1992). Final Report. Deliverable 30. McDonald et. al. DGXIII, Brussels, Belgium.

Imacho, N., Ichinose, T. and Yoshikai, T. (1995). The Verification of the Guidelight Effectiveness
and the System of the Future. Proc. of the 2nd World Congress on Intelligent Transport Systems,
pp 1152-6. VERTIS.

Inoue, N., Iida, Y. and Hasegawa, T (1990). On-Ramp Control of Freeway Network. Concise
Encyclopaedia of Traffic & Transportation Systems (Editor: Papageorgiou, M.), Pergamon press,
1990. pp 300-3.

INVAIDII. (1994). Final Report on the Field trial in Antwerp. Deliverable 13. Blosseville J.M.,
Morin J.M., Guillen S., Chrisoulakis J., Cypers L. Renard M. DGXIII, Brussels, Belgium.

INVAIDII. (1995). Final Report of the Field Trial in urban sites: data collection and evaluation.
Deliverable 14. Ahsbahs C., Boillot F., Lenoir F., and Boulmakoul A., et. al. DGXIII, Brussels,

Janssen, W., Kuiken, M. and Verwey, W. (1994). Evaluation Studies of a Prototype Intelligent
Vehicle. Proc. of the 1st World Congress on Applications of Transport Telematics and IVHS, pp
2063-70. ERTICO.

Kajiya, Y., Fukuzawa, Y., and Ishimoto, K. (1995). Development of Intelligent Delineator System.
Proc. of the 2nd World Congress on Intelligent Transport Systems, pp 1193-8. VERTIS.

DIATS                                     Deliverable 1

150796                                                                       DIATS/TRG/WP100

Kerner, B.S. and Konhauser, P. (1994). Structure and Parameters of Clusters in Traffic Flow. Phys.
Rev. E, 50, 1, pp 54-83.

Komoda, N. et. al. (1995). The Latest Major ITS Activities in Japan. Overview, Trends and Future
Scope. Proc. of the 2nd World Congress on Intelligent Transport Systems, pp 2589-94. VERTIS.

Krause, G. (1995). COMPANION: New modules for Traffic Guidance Technology. Traffic
Technology International, pp. 210-212, Surrey, UK.

Krause, G. (1996). Looking for a Roadside COMPANION? Traffic Technology International
(Quarterly), Feb/Mar., p.4, Surrey, UK.

Leasure Jnr., W. A. and Burgett, A. L. (1994). NHTSA's IVHS Collision Avoidance Research
Program: Strategic Plan and Status Update. Proc. of the 1st World Congress on Applications of
Transport Telematics and IVHS, pp 1916-23. ERTICO.

McMurran, R. (1994). The Integration of Driver Information Systems. Traffic Technology
International 94. pp 122-4.
Middelham, F. (1994). State of the Art in Dynamic Traffic Management in The Netherlands. 7
IFAC/IFORS/IFIP Symposium on Transportation Systems, Tianjin, China.

Miyazaki, T. (1995). Promotion of the Program of Advanced Safety Vehicles for the 21st Century.
Proc. of the 2nd World Congress on Intelligent Transport Systems, pp 2612-17. VERTIS.

Mizoguchi, M. (1995). VICS - Strategy and Deployment Plan. Proc. of the 2nd World Congress
on Intelligent Transport Systems, pp 2618-21. VERTIS.

Morello, E., Benz, T. and Ludman, J. (1994). AICC Assessment. Proc. of the 1st World Congress
on Applications of Transport Telematics and IVHS, pp 1900-07. ERTICO.

Morin, J.M. (1981). Etude du feu optimal en un accès controlé. IRT report.

Morin, J-M. (1982). La regulation de vitesse sur autoroute en France et a l'etranger (Speed control
on motorways in France and other countries). IRT Research Report No 56.

Morin, J. M. (1996). La détection automatique d'incidents sur autoroute. Synthesis Report for
USAP (Union des Société d'Autoroute à péage). Isis Consultant.

Naab, K. and Reichert, G. (1994). Driver Assistance for lateral and longitudinal Vehicle Control.
Proc. of the Int. Symp. on Advanced Vehicle Control. Tsukuba, Japn. pp 449-54.

Nakamura, M. and Yoshikai, T. (1995). The Experiments on Lateral and Longitudinal Control
System in ARTS. Proc. of the 2nd World Congress on Intelligent Transport Systems, pp 1069-74.

DIATS                                     Deliverable 1

150796                                                                          DIATS/TRG/WP100

Owens, D. Schofield, M.J. (1988). Access control on the M6 Motorway; evaluation of Britain's
first ramp metering scheme. Traffic Engineering & Control, Dec 1988.

Panizza, E. and Levizzari, G. (1990). Development Underway in Fiat for On-board Information
and Collision Avoidance Systems. Vehicle Electronics in the 90's . ISBN 1-56091-047-x.

Papageorgiou M, Hajsalem H, Blosseville, J.M. (1989). ALINEA A local feedback control law for
on-ramp metering 70th Annual Meeting Transportation Research Board, Paper n° 910659 TRB,
Washington D.C

Papageorgiou, M. (1990). On-ramp control - Co-ordinated Time of Day Strategies. Concise
Encyclopaedia of Traffic & Transportation Systems (Editor: Papageorgiou, M.), Pergamon press,
1990. pp 285-94.

Parent, M. and Fanconnier, C. (1995). Design of an Electric Vehicle Specific for Urban Transport.
EVT 95.

Prynn, J. (1995). New cat's eye waits for green light. The Times, 'CAR' supplement, p12. April 8th,

Quinlan, T. F. and Gouse, S. W. (1995). Planning the 1997 Proof of Technical Feasibility
Demonstration for the United States Automated Highway System Program. Proc. of the 2nd
World Congress on Intelligent Transport Systems, pp 2514-19. VERTIS.

Rao, B.S.Y. and Varaiya, P. (1994). Flow Benefits of Intelligent Autonomous Cruise Control in
Mixed Manual and Automated Traffic. Transportation Research Record, 1408, pp 36-43.

Ross, P. (1990). On-Ramp Control - Realisation Principles. Concise Encyclopaedia of Traffic &
Transportation Systems (Editor: Papageorgiou, M.), Pergamon press, 1990. pp 303-6.

Roussel, J.C, Petrucci, J., Borie, B., Gautier, Y. (1995). Automatic Incident Detection, SAPRR
test Program. Revue Générale des Routes, n° 735, Décembre 1995, pp 25-32.

SACTRA (1994). Trunk Roads and Generation of Traffic. HMSO, U.K.

Sala, G. and Pressi, F. (1994). Strategies for Intelligent Cruise Control Application. Proc. of the 1st
World Congress on Applications of Transport Telematics and IVHS, pp 2039-46. ERTICO.

SAMOVAR (1995). Edited Final Report. Deliverable D11. Ed. Fincham et. al. DGXIII, Brussels,

Satoh, H. and Taniguchi, I. (1994). A New Control Method of Adaptive Cruise Control. Proc. of
the 1st World Congress on Applications of Transport Telematics and IVHS, pp 2094-2101.

DIATS                                       Deliverable 1

150796                                                                        DIATS/TRG/WP100

Saxton, L. G. and Bishop, J. R. (1994). The United States Department of Transportation
Automated Highway System (AHS) Program. Proc. of the 1st World Congress on Applications of
Transport Telematics and IVHS, pp 1952-8. ERTICO.

Schumacher, R. W., Olney, R. D., Wragg, R., and Landau, F. H., et. al. (1995). Collision Warning
System Technology. Proc. of the 2nd World Congress on Intelligent Transport Systems, pp 1138-

Shladover, S. (1978). Dynamic Entrainment of Autonomous Guided Transit Vehicles. High Speed
Transportation Journal, 12, 3, pp 87-113.

Shladover, S. E. (1995). Recent Steps Forward in PATH Research On Advanced Vehicle Control
Systems (AVCS). Proc. of the 2nd World Congress on Intelligent Transport Systems. VERTIS.

Smulders, S. (1992). Control by variable speed signs: The Dutch Experiment. Ministry of
Transport, The Netherlands.

SPECTRUM (1993). Field Trial Assessment Report. Deliverable 15. DGXIII, Brussels, Belgium.

Staehl, A., Oxley, P., Berntman, M. and Lind, L. (1994). The Use of Vision Enhancements to
Assist Elderly Drivers. Proc. of the 1st World Congress on Applications of Transport Telematics
and IVHS, pp 1999-2007. ERTICO.
Stephanedes, Y.J. and Chassiakos, A.P. (1991). A low-pass filter for incident detection. Proc., 2
Int. Conf. Applic. Adv. in Transp. Engrg., ASCE, New-York, N. Y., pp 378-82.

Stephanedes, Y.J. and Chassiakos, A.P. (1993). Freeway incident detection through filtering.
Transportation Research C, 1, 3, pp 219-33.

Stephanedes Y.J. and Chassiakos A.P. (1993). Application of filtering techniques for incident
detection. ASCE, Journal of Transportation Engineering, no 119 pp 13-26.

Taale, H.and Van Velem, G.A. (1996). The assessment of multiple ramp-metering on the ringroad
of Amsterdam. 8 Conference on Road Traffic Monitoring & Control. IEE, Conference publication
n° 422, pp 164-8.

Taniguchi, E, Tsuge, A. and Hayama, A. (1994). Traffic Impediment Monitoring/Warning Systems
by Image Processing. Proc. of the 1st World Congress on Applications of Transport Telematics
and IVHS, pp 1933-40. ERTICO.

Tomioka, Y., Sugita, E. and Gonmori, A. (1995). The Development of the Rear End Collision
Warning System Equipped with a Radar for Heavy Duty Trucks. Proc. of the 2nd World Congress
on Intelligent Transport Systems, pp 1134-7. VERTIS.

Transport Technology Publishing Ltd. (1996). The International ITS Index. Transport Technology
Publishing Ltd., Binghampton, NY, USA.

Tribe, R., Prynne, K., Westwood, I. and Clarke, N. et. al. (1995). Intelligent Driver Support. Proc.

DIATS                                      Deliverable 1

150796                                                                     DIATS/TRG/WP100

of the 2nd World Congress on Intelligent Transport Systems, pp 1187-92. VERTIS.

TRITEL. (1994). Control Schemes for Speed Harmonisation and Queue Prevention. TRITEL,
Transport Infrastructure and Telematics.

Tsugawa, S. and Fujii, H. (1995). ITS Activities of Japanese MITI: AVCS Technologies in Focus.
Proc. of the 2nd World Congress on Intelligent Transport Systems, pp 2606-11. VERTIS.

U.S. Dept. of Transport and F.H.W.A. (1996). Intelligent Transport Systems Projects. F.H.W.A,
Washington D.C., U.S.A.

Verwey, W. B. and Kaptein, N. A. (1994). Automatic Man-Machine Interface Adaptation to
Driver Workload. Proc. of the 1st World Congress on Applications of Transport Telematics and
IVHS, pp1701-08. ERTICO.

Virginia Transportation Research Council. (1994). Predicting Travel Volumes for HOV Priority
Techniques - User's Guide. U.S. FHWA. VTRC 94-Rl3.

Watanabe, T., Kishimoto, N., Hayafune, K, and Yamada, K. et. al. (1995). Development of an
Intelligent Cruise Control System. Proc. of the 2nd World Congress on Intelligent Transport
Systems, pp 1229-35. VERTIS.

Yang, H. Yagar, S., (1994). Traffic Assignment and Traffic Control in General Freeway arterial
corridor systems. Transportation Research B, 28B, n°6, pp 463-86.

Yang, H. Yagar, S., IIda, Y. and Asakura, Y. (1994). An algorithm for the inflow control problem
on urban freeway networks with user-optimal flows, Transportation Research B, 28B, n°2, pp 123-

Zhang, X. (1992). Intelligent Driving - PROMETHEUS Approaches to Longitudinal Traffic
Control. Proc. of the Intelligent Vehicles Symposium, Detroit, IEEE.

DIATS                                    Deliverable 1

150796                                                                            DIATS/TRG/WP100

ANNEX A:           CORD Recommendations           of    Transport    Telematics     Functions     and

The following definitions of transport telematics functions and subfunctions are relevant extracts
taken from CORD Deliverable D004 - Part 3 (CORD 1994), and is included here for completeness.


Users: Traffic Authorities

Subject: Measures for traffic control in conformity with demand management principles.

F3.1     Section Traffic Control

Includes all measures designed to control the flow within specific links of network.

         SF3.1.1        Section state monitoring
                        Obtain data on traffic status or characteristics, pre-processes data for flow
                        estimation, and monitors congestion.

         SF3.1.2        Incident detection/accident detection and identification
                        Detects and identifies abnormal occurrences relevant for the state of traffic.

         SF3.1.3        Section control computation
                        Computes, on the basis of control strategies, control measures such as
                        speed limitations and recommendations, lane allocation ramp metering etc.

         SF3.1.4        Section control Actuating
                        Activates control devices, either individual or public.

         SF3.1.5        Local speed enforcement
                        Activates control measures on road sections of high potential risk.

F3.2     Intersection Traffic Control

         SF3.2.1        Intersection state monitoring

         SF3.2.2        Intersection control computation
                        Effects control based on the integration of empirical data and model based

         SF3.2.3        Intersection control actuating
                        Operates control devices, eg traffic lights for ramp metering, and variable
                        message signs.

DIATS                                       Deliverable 1

150796                                                                          DIATS/TRG/WP100

F3.4     Localised Area Traffic Control

Applies to specific locations or areas in the network, which require dedicated strategies due to the
specific operational environment.

         SF3.4.1         Tidal Flow Control

         SF3.4.2         Ramp Control
                         Implements control measures for merging traffic flows at motorway
                         intersections, taking into account actual traffic demand and the actual
                         capacity of the road section ahead. The measures are based on calculations
                         using traffic monitoring functions and devices, such as ramp, metering.

         SF3.4.3         Tunnel Traffic Control

         SF3.4.4         Bridge Traffic Control

         SF3.4.5         Lane Management
                         Implements control measures reserving certain traffic lanes exclusively to
                         specific classes of vehicles, eg. high occupancy vehicles or buses.


Users: Vehicle Drivers

Subject: This family of functions embraces the monitoring of drivers, vehicles and surroundings,
and provides indirect or direct assistance for driving.

F9.1     Monitoring Environment and Road

Acquires information on the current status of the immediate environment of the vehicle.

         SF9.1.1         Road surface and marking monitoring
                         No combination with meteo data is intended.

         SF9.1.2         Road geometry monitoring
                         Monitors the spatial dimension of the road ahead of the vehicle to establish
                         manoeuvring restrictions.

         SF9.1.3         Road visibility monitoring.

         SF9.1.4         Road regulations monitoring
                         Monitors roadside information such as traffic signals and signs.

F9.2     Monitoring Driver

Observes the driver's control of the vehicle, the drivers psychological condition and evaluates a
possible deviation by him/her from normal (safe) behaviour.

DIATS                                       Deliverable 1

150796                                                                           DIATS/TRG/WP100

         SF9.2.1       Driver status monitoring

         SF9.2.2       Creation of driver profile and identification of trends

F9.3     Monitoring Vehicle

Acquires, processes and records data on vehicle dynamics and operational status to diagnose and
predict vehicle failures and vehicle dynamic behaviour.

         SF9.3.1       Vehicle dynamics monitoring
                       (eg. speed).

         SF9.3.2       Vehicle operational status monitoring
                       (eg. identification of current vehicle parameters, degradation or failure).

         SF9.3.3       Vehicle status recording
                       Records vehicle (and driver?) data of relevance for accident or performance

F9.4 Vision Enhancement
Improves the visibility of the driving scene by autonomous and non-co-operative means, in sub-
normal visibility conditions by providing direct visual information to the driver.

F9.5 Collision Risk Estimation
Detects potential obstacles in relation to the dynamics and predicted trajectory of a moving vehicle
for collision avoidance.

         SF9.5.1       Relative position determination

         SF9.5.2       Conflict zone monitoring and trajectory prediction
                       Detects and monitors the dynamic status of other road users within the
                       manoeuvring zone of the vehicle in order to identify trajectories that may
                       lead to conflicts.

         SF9.5.3       Safety margin determination
                       Continuous determination of the range of vehicle performance within which
                       stable driving manoeuvres can be performed, considering all relevant
                       influencing factors related to the vehicle and the environment, both in actual
                       and predictive manner. The steps involved are determination of maximum
                       potential limits; determination of the actual and predicted position of the
                       driving situation; and determination of safety margins.

         SF9.5.4       Critical course determination
                       Continuous determination of a safe trajectory with regard to road
                       boundaries, stationary and moving objects. This consists of determination
                       of own trajectory; identification of potential collisions with obstacles; and a
                       course which minimises the risk of collision.

DIATS                                      Deliverable 1

150796                                                                           DIATS/TRG/WP100

F9.6     Actuator Control (Dynamic Vehicle Control)

Control of lateral, longitudinal and vertical dynamic behaviour of the vehicle to influence the vehicle
in order to stabilise driving according to the demand of the driver or higher level control system.
Actuators are controlled in accordance with the control strategy for the actual situation:this ranges
from advice to the driver on supportive actions to fully automated control.

         SF9.6.1        Lateral activation/control

         SF9.6.2        Longitudinal activation/control

F9.7     Dialogue Management

Manages the information flow from the vehicle/driver, and dynamically optimises the man-machine
interface (MMI) regarding priority and sequence of information presented to or derived from the
driver through any type of interface.

         SF9.7.1        User presentation
                        Identifies priorities of incoming information, optimises and integrates
                        information or actual user interface.

         SF9.7.2        Driver tutoring
                        Provides feedback to the driver regarding non compliance with traffic rules,
                        regulations and a safe dynamic driving status.

         SF9.7.3        Information requests
                        Provides the driver with suitable input facilities for requests for information,
                        eg route guidance, parking.

         SF9.7.4        Facility requests
                        Provides the driver with suitable input facilities for requests for making

         SF9.7.5        Traffic condition reporting
                        Provides the driver with suitable input facilities for reporting

         SF9.7.6        Automatic reporting
                        Provides automatic reports to infrastructure, eg link journey times.

DIATS                                       Deliverable 1

150796                                                                 DIATS/TRG/WP100

ANNEX B:      Selected Glossary of Abbreviations

AC-ASSIST           Anti-Collision Autonomous Support and Safety Intervention
                    SysTem/Research into Optimally Assisted Driving Systems and
                    Technology developed for European Roads. 4th Framework Project,
ADR                 Accident Data Recorder
ADVANCE             Advanced Driver and Vehicle Advisory Navigation Concept
AHS                 Automated Highway System
AICC                Autonomous Intelligent Cruise Control
AID                 Automatic Incident Detection
ALINEA              Asservissement Lineaire des Entrees d'Autoroute
ARIADNE             Application of Real Time Intelligent Aid for Driving and Navigation
                    Enhancement. DGXIII DRIVE2 Project, V2004.
ARTS                Advanced Road Transportation System
ATIS                Advanced Travellers Information Systems
ATMS          Advanced Traffic Management Systems
ATT                 Advanced Transport Telematics

CAS              Collision Avoidance system
CEC              Commission for the European Communities
CGEA       Compagnie Generale d'Entreprises Automobiles
CHRISTIANE Motorway Traffic Flow Monitoring and Control. DGXIII DRIVE1 Project,
COFIROUTE COmpagnie Financiere et Industrielle des autoROUTEs
CORD       Co-Ordination project for Research and Development. DGXIII DRIVE2 Project,
CSST             Centro Studi Sui Sistemi di Transporto
CW               Collision Warning

D'ACCORD            Development and Application of Co-ordinated Control of Corridors. 4th
                    Framework Project, TR1017.
DACD         Driver Assistance and Cooperative Driving. A cross project collaborative study
                    conducted as part of the CORD project.
DETER               Detection Enforcement and Tutoring for Error Reduction. DGXIII
                    DRIVE2 Project, V2009.
DIM                 Driver Impairment Monitoring
DRACO               DRiver and Accident Coordinating Observer. DGXIII DRIVE Project,
DREAM               A Feasibility Study for Monitoring Driver Status. DGXIII DRIVE1
                    Project, V1004.
DRIVE1 and 2 Dedicated Road Infrastructure for Vehicle Safety in Europe. The DRIVE1
                    program lasted from 1988-91, and its successor, DRIVE2 from 1991-94.

EDDIT         Elderly and Disabled Drivers and Information Telematics. DGXIII DRIVE2
                     Project, V2031.
EDF                  Electricite De France

DIATS                                  Deliverable 1

150796                                                                    DIATS/TRG/WP100

EMMIS               Evaluation of Man Machine Interface by Simulation. DGXIII DRIVE2
                    Project, V2006.
ERTICO              European Road Transport Telematics Implementation Coordination
EU                  European Union
EUREKA              EUropean REsearch Coordination Agency
EUROCOR             EUROpean urban CORridor control. DGXIII DRIVE2 Project, V2017.

FHWA         Federal Highway Administration
FOCAS                Fostering the development evaluation, and deployment of FOrward Crash
                     Avoidance Systems.

GIDS                GenerIc Driver Support systems. DGXIII DRIVE1 Project, V1041.

HARDIE              Harmonisation of ATT Roadside and Driver Information in Europe.
                    DGXIII DRIVE2 Project, V2008.
HERMES              High Efficiency Roads with Rerouting Methods and Traffic Signal Control.
                    DGXIII DRIVE2 Project, V2019.
HF                  Haptic Feedback
HGV                 Heavy Goods Vehicle
HOPES               HOrizontal Project for the Evaluation of Safety. DGXIII DRIVE2 Project,
HOV                 High Occupancy Vehicle

ICARUS               Interurban Control And Road Utilisation Simulation. DGXIII, DRIVE1
                     Project, V1052
ICC                  Intelligent Cruise Control
INRESPONSE Incident RESPonse with ON-line Innovative SEnsing. 4th Framework Project,
INRETS               Institut National de Recherche sur les Transports et leur Securite
INRIA       Institut National de Recherche en Informatique et Automatique
INVAID               Integration of Video for Automatic Incident Detection
ITS America          Intelligent Transportation Systems - America

LIDAR        Laser-Radar. A rangefinding system using the radar principle but using a laser beam
                    to effect detection.

MELYSSA             Mediterranean-Lyon-Stuttgart Site for ATT. DGXIII, DRIVE2 Project,
METALINE            METALINE is made from META and LINE (LINEARISED) because the
                    speed equation of the simulation model META is used (after linearisation)
                    in METALINE
MMI                 Man-Machine Interface
MOTIV               Mobilitat und Transport im Intermodalen Verkehr (Mobility and Transport
                    in Multi-modal Traffic).

ODIN                Optical Distance measurement INnovation

DIATS                                   Deliverable 1

150796                                                         DIATS/TRG/WP100

PATH         Partners for Advanced Transit and Highways
PRAXITELE    Public pRAtique, eXperimental, Individuel Transport ELEtrique
PREDIT       Programme de REcherche et de Developpement pour l'Innovation et la
             Technologie dans les transports terrestres
PRO-GEN      PRO-GENERAL - a sub programme of the EUREKA PROMETHEUS
PROMETHEUS   (EUREKA) PROgraMme for a European Traffic system with Highest
             Efficiency an Unprecedented Safety
PROMOTE      PROgramme for MObility and Transportation in Europe
PSA          Precursor Systems Analysis

RIDIS        Road Incident Detection and Information Service
RTA          Road Traffic Advisor

SAMOVAR      Safety Assessment Monitoring On-Vehicle with Automatic Recording.
             DGXIII DRIVE2 Project, V2007.
SAVE         System for effective Assessment of driver state and Vehicle control in
             Emergency Situations. 4th Framework Project, TR1047.
SOCRATES     System Of Cellular RAdio for Traffic Efficiency and Safety. DGXIII
             DRIVE2 Project, V2013.
SPECTRUM     Strategies for the Prevention of Road Trafiic Congestion. DGXIII DRIVE1
             Project, V1059.

TABASCO      Telematics Applications in Bavaria, Scotland and Others. 4th Framework
             Project, TR1054.
TESCO        Test On Cooperative Driving. DGXIII DRIVE2 Project, V2010.
TRL          Transport Research Laboratory (UK). Now under private ownership as the
             TRF, Transport Research Foundation.

UDC          Urban Drive Control. 4th Framework Project, TR1060.
UMTRI        University of Michigan Transport Research Institute

VERTIS       VEhicle Road and Traffic Intelligence Society
VMS          Variable Message Sign

WP           Work Package

DIATS                          Deliverable 1


To top