1. DESCRIPTIVE TITLE: Transportable Manned and Robotic Digital Geophysical Mapping
(DGM) Tow Vehicle
2. ESTCP THRUST AREA: 1) Unexploded Ordnance (UXO) Detection, Discrimination, and
3. LEAD ORGANIZATION: U.S. Army Corps of Engineers Engineering & Support Center-Huntsville
(USAESCH), PO Box 1600, Huntsville, Alabama 35807-4301; Scott Millhouse, PE, ED-CS-D phone
256-895-1607 fax 256-895-1602 email: firstname.lastname@example.org
a.) Objective: This project will integrate an innovative robotic tow vehicle with industry standard Digital
Geophysical Mapping (DGN) sensors and advanced geo-location positioning equipment to autonomously
map target demonstration areas. Phase I will focus on integration and path following to precisely
replicate target coverage for multiple runs with multiple sensors. Phase II will continue on phase I
developments and focus on more challenging site conditions that require obstacle recognition and
avoidance and a secondary positioning system to maintain path when the primary positioning system is
b.) Technology Description: The principle DGM tow vehicle is the Segway Robotic Mobility Platform
(RMP) in both a two wheel RMP 200 and 4 wheel RMP 400 ATV configurations. This vehicle is battery
powered and can be quickly broken down for express package shipment like the typical DGM sensors.
Phase I positioning is provided to sub-centimeter accuracy by the ArcSecond Laser “Indoor GPS” triad
package. Phase II as envisioned will utilize DGPS with INS, electronic magnetic compass, and dead
reckoning solutions for positioning and a form of synthetic vision using 2-D LIDAR for obstacle
recognition. Phase II will be demonstrated in autonomous, semi-autonomous and tele-operated modes.
c.) Expected Benefits: The robotic solution will permit more precise path following than by man towed
equipment. The immediate benefit for sensor development and demonstration/prove out sites is to cover
the exact same pathway multiple times with changed parameters to the same or with alternate sensors. For
sites with explosive or safety concerns the robotic DGM vehicle will eliminate personnel risk. Operation
will be monitored remotely. The geophysicist will be no longer be a “beast of burden” towing equipment.
He can focus his efforts to monitor and acquire the highest possible quality data. It is expected that not
only will quality dramatically improve but also so will productivity.
5. Problem Statement:
Unexploded Ordnance (UXO) poses a threat to both human life and the environment. Millions of
UXO are located in the U.S. on active test and training ranges and Formerly Utilized Defense Sites
(FUDS). In addition to the millions of UXO, there are many times more cultural and debris anomalies.
Digital Geophysical Mapping (DGM) is used to map the areas and to locate, identify and select the items
for sampling and removal. Many modes of DGM are utilized that include airborne and ground based
platforms such as large towed arrays and man-portable equipment. All sites will need some amount of
ground based mapping by man-portable or narrow width towed platforms depending upon terrain, ground
cover and UXO objective. Man portable equipment essentially uses the operator as a “beast of burden” to
carry the electronics and batteries and tow or swing the sensors in addition to monitoring the equipment
and maintaining track. This can lead to reduced data quality due to deviation from pathway and
inadequate sensor and position monitoring. Fatigue can cause reduced production and inattention to safety
concerns. This proposal will initially test and develop an easily transportable battery powered tow vehicle
that has been measured and documented for its effect on typical geophysical sensors.
New sensors, analysis techniques and field methodology are normally selected by a geophysical
prove out process. Typically an area is seeded with the full range of expected Ordnance and Explosive
items and clutter at a range of depths as a site specific evaluation or as part of a technology validation
such as at the Yuma or Aberdeen Proving Grounds Standard UXO Technology Demonstration Sites. For
the best comparisons the individual sensors should cover the same pathways as close as possible.
Analysis algorithms are predicted to require a position accuracy of better than +/- 2 cm in x, y and z
position to be effective. The next step is a robotic enhanced tow vehicle that will utilize the ArcSecond
Indoor GPS positioning system to maintain the position accuracy and accurately maintain pathway track.
All sensor suites being considered in evaluations will traverse the site by being robotically towed to the
same pathway. This will facilitate analysis algorithm development and equipment performance
Typical field mapping areas have varied terrain, obstacles, line-of-sight and horizon visibility.
This adds challenges to maintaining a pathway and for maintaining sensor positioning. Semi-autonomous,
autonomous and tele-operated mapping operations are desired for enhanced quality, personnel safety and
production. Travel pathways can be designed to minimize interferences and for maximum production
with the objective to survey all areas that are currently being covered by man portable towed sensors. For
the most flexibility and production, autonomous obstacle recognition and avoidance and a methodology
for maintaining track and positioning will be required for areas with the primary geo-location positioning
system is shadowed.
6. Technology Description
a) Technology objectives. We propose to develop the Segway® Human Transporter (HT), Cross-Terrain
Transporter (XT) and the two wheeled and four wheeled versions of the Robotic Mobility Platform
(RMP), RMP 200 and 400, as a man powered DGM and robotic narrow towed array DGM tow vehicle.
The equipment is planned for full performance testing as a manned tow vehicle to define capabilities and
limitations and then augmented for two and four wheeled operation for tele-operated, semi-autonomous
and autonomous robotic operation. Our objective is a small, economic, environmentally clean easily
shippable tow vehicle. Phase II will add capability for more autonomous operation for larger less
controlled locations by including obstacle recognition and avoidance and positioning and path following
for when the primary positioning system is shadowed. Phase II will provide the capability to survey all
areas that are typically surveyed by man portable towed sensors.
b) Technology description. The Segway® Cross-Terrain Transporter (XT) is the latest self-balancing
human transporter that provides enhanced performance on a variety of terrain with minimal
environmental impact. Featuring all-terrain tires, new extended-range lithium-ion batteries and specially
tuned software, this rugged Segway XT will go practically anywhere you want to go.
The Segway XT has been specifically developed for stability, comfort and performance on uneven and
tough terrain. The Segway XT's low-pressure tires smooth the ride on bumpy surfaces and minimize trail
impact, while the wider track increases stability on uneven ground. The Segway XT's software has been
modified to support the new tire size and to supply improved control and performance. The increased
energy capacity of the standard lithium-ion batteries support the demands of the Segway XT, while still
providing a 10 mile off-pavement range, depending on terrain, riding style and payload.
Segway® Human Transporter (HT) can self-balance because of a technology called dynamic
stabilization. Dynamic stabilization works in much the same way your own sense of balance does. Where
you have an inner ear, eyes, muscles, and a brain to keep you balanced, the Segway HT has solid-state
gyroscopes, tilt sensors, high-speed microprocessors, and powerful electric motors performing to keep it
balanced. Working in concert, these extensively tested, redundant systems sense your center of gravity,
instantaneously assess the information, and make minute adjustments one hundred times a second.
Segway HT balances whether you're traveling at 10 mph, carrying a heavy load, slowly maneuvering in
tight spaces, or standing perfectly still. See Appendix B for additional details on the technology behind
The Robotic Mobility Platform (RMP) version of the HT is available in the typical 2 wheel mode as well
as a dual unit with 4 wheels. The 4 wheel version provides the additional stability and traction for rough
terrain applications of 4 individually powered wheels. This platform is based upon the mainstream
commercial product except that it provides output to a generic PC external command control computer
and takes instruction from that computer for maneuvering instead of relying upon the commercial
product’s manual control input. Instead of having a balancing rider providing direction by leaning or turn
commands by the operation of a twist grip, the command computer communicates with the HT by the
CANbus interface at 100Hz. The unit comes with a standard joystick mode for remote operation as well
as a command set for the unit and the decode for the state variables as sent by the RMP to the computer
for robotic control. See Appendix B for additional details on the RMP.
The proposed Segway solutions meet our objectives for a small, economic, environmentally clean easily
shippable tow vehicle for DGM.
c) Technology Maturity. The HT has been commercially available since March 2002 with many thousand
sold and supported worldwide. The Robotic Mobility Platform (RMP) has been available since 2003 and
is part of an active DARPA program with 12 University partners to develop to support military missions.
The RMP is sold primarily to Labs, Universities and Government for robotic development with over 45
d) Technical Approach. We propose the following program:
I. Pre-project- Segway HT testing (Self performed by the Huntsville Center at no cost to ESTCP -
A current-off-the-shelf used refurnished HT was purchased by USAESCH and tested for the
static and the dynamic effect of the ferrous and non-ferrous components and for production of
Electromagnetic (EM) fields.
First an area was established that was clear of subsurface anomalies. Then an EM-61 was set up
and zeroed to background with a non-metallic tape stretched out from the center of the coil. The
HT was then moved powered off towards the EM with EM measurements taken at 1’ intervals
from 20’ to 2’ from the coil center. The test was done several times with the equipment powered
off. The test was then performed several times once again with the HT manned and powered up
and in balanced mode. Measurements were observed during movement towards the geophysical
sensor with readings recorded at the 1’ intervals. Runs were consistent with the surprise of no
difference between the HT powered on or off. As shown on the plot the unit did not affect the EM
at approximately 2 meters from the center of the coil.
Testing of the affects of the
Segway vehicle to an EM-61
Walk-away Test Results
Geonics mV (Std 61)
18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2
Feet from Center of 1M X 1M Coil
Average - top coil reading (mV) Average - bottom coil reading (mV)
The test was repeated in similar fashion using the G-858 magnetic sensor in gradiometer mode.
For these tests the HT strongly affected the gradiometer at approximately 1 meter and tapered to
background at approximately 2 meters in the manned dynamic mode.
Testing of the affects of the Segway
vehicle to an G-858 magnetometer
in gradiometer mode
Geometrics 858 gradiometer & SEGWAY Walk-Away Test
18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1
Feet from center of magnetic sensor
Power On - Avg Gradient (nT/M) Power off - Avg Gradient (nT/M)
Based upon these results a tow bar length to get a minimum of 2 meters separation from the rear
of the HT to the center of a sensor is required. The tow bar yoke is designed to be fully
articulating so that no differential movement is transmitted between the senor and tow vehicle. A
non-metallic wheel is provided for terrain following and balancing with tow bar connection
allowing vertical and horizontal articulation. All components used to build the tow system are
wood, composite or plastic so they do not affect any towed sensor.
Composite articulating towbar and hitch
This device was tested for suitability as a tow vehicle with typical EM-61 equipment at the
USAESCH Redstone Arsenal McKinley Test Site. This site is a seeded open grassed field. The
test site had not been mowed in months so the grass was approximately a foot high. In addition
the unit’s battery was slightly depleted from being off charge for several days. Regardless it still
provided approximately 3 hours of operation. The integrated system included the EM-61 and tow
vehicle but not a primary positioning system due to non-availability (all at APG). Fiducial
measurements were established and the array traversed typical DGM mapping lanes. It was easily
controlled, highly maneuverable with adequate power and traction in grass even with the standard
smooth pavement tires. The smooth tires were inadequate for the sand pit test area where the loss
of traction caused the Segway to immediately shut down. Visualization and analysis of the
acquired data set has shown no difference between the results from a man towed system even
including the fiducial lag shown in the included plot. These efforts were performed at no cost to
ESTCP by USAESCH. ($20k contribution by USAESCH)
Segway tow vehicle with EM-61
electronics and data pack/recorder
Segway tow vehicle with EM-61
electronics and data pack/recorder
at McKinley Range
II. Phase I- Segway Robotic Mobility Platform Application
Based upon the pre-project findings a Segway XT and RMP200 and 400 will be procured and
modified to minimize the effect to the DGM sensors and to optimize it as a tow vehicle. The XT
and RMP vehicles will be initially developed and tested as the two wheeled version in a similar
fashion to the pre-project testing by operation in the remotely operated joy stick mode. The
testing will then be repeated and the performance evaluated as configured in a four wheeled
version. It is envisioned that the two wheel mode will be adequate for areas with smooth terrain
and low vegetation with a light payload. The four wheel version may be required for the more
challenging conditions. We will define the limits of terrain and payload for the two
The system will then be augmented by the addition of the ArcSecond Indoor GPS laser based
positioning system and by a path following robotic control program on a PC control computer.
For the geophysical sensor positioning will be provided by the ArcSecond Triad positioning
system uses a rover sensor array that senses x-y-z position and multi-axis movements to sub-
centimeter accuracy at a 40 Hz refresh rate for up to a 5 acre area. Since the tow vehicle and
geophysical sensor’s relative position changes with terrain and maneuvering a separate position
sensor will provide the tow vehicle location. The software will compare the tow vehicles and
geophysical sensor’s positions to the desired path to compute the lever arm and then instruct the
tow vehicle to maintain the geophysical sensor on track. This accurate position with a high
refresh when integrated with the Segway 100Hz control system to position and maintain the
DGM vehicle along the planned pathway will result in an unprecedented accuracy and adjustment
speed for path following. The principle advantage with this concept is that a site (geophysical
prove out or technology demonstration) can be traversed many times with multiple sensors
configurations traveling at exactly the same design pathway and speed over the subsurface
anomalies. This should provide laboratory test results quality in a real field applications.
1. The system will be repetitively tested at the McKinley Range to validate
performance and data quality using EM and magnetometer DGM sensors and the
various tow vehicle configurations.
2. Upon successfully completing the McKinley testing, the system suite will be
demonstrated at the APG Standard UXO Demonstration Site’s Calibration Grid
using EM and magnetometer DGM sensors. The acquired data sets will be
provided to researchers for in depth analysis using the newly developed
predictive models. (Funded outside this project by existing and USAESCH
3. Results will be documented into a Phase I report, presented at the IPR and the
SERDP/ESTCP Environmental Conference for a Go/No go decision on Phase II.
Phase I Technical Details
The Phase I effort will be focused on demonstrating capability of both teleoperation of the
Segway tow vehicle/DGM trailer system and autonomous path following in an obstacle-free
environment. Work will be divided into two main tasks: a) video feedback/ manual teleoperation,
followed by b) autonomous navigation and path following. The communication and control
system hardware will include two laptop computers (one mounted on the Segway tow vehicle, the
second remotely located and functioning as the user interface). The two computers will
Video feedback and manual teleoperator control
1. A low cost video camera (InsideOut Watchport V) mounted on a position controllable
platform (Eagletron Trackerpod) will be installed on the tow vehicle. The Trackerpod
platform and associated control software gives 160˚ pan and 110˚ tilt of the mounted camera,
and connects to the tow vehicle computer's USB port (powered USB port is preferred). The
camera is also connected to the computer through a USB port.
2. The control commands for the Segway RMP will be entered through a graphical user
interface (GUI) implemented on the laptop computer. The GUI will be implemented using
Microsoft Visual Basic, and will run simultaneously with the video camera software on the
Windows XP operating system.
3. To achieve teleoperation, the laptop computer on the Segway will need to be remotely
accessble. A JAVA-based remote access and computer desktop control service and software
system will be used to connect the Segway computer and remote computer over the Internet
(for example: http://www.logmein.com).
The teleoperator system computers will communicate video and command signals over an
802.11b/g wireless network. The computer on the tow vehicle and the remote computer will be
equipped with commercially available wireless network interface cards. If necessary, physical
range can be extended by additional radio frequency amplifiers.
1. A video feedback/teleoperation and wireless communication system similar to that described
above has been implemented and improved over two generations of remotely piloted vehicles
at Auburn University. Those vehicles were designed to study the efficacy of low-cost remote
control over Internet protocol networks. For the proposed tow vehicle project, a dedicated,
closed peer-to-peer wireless network will also be explored, so that extraneous network traffic
2. The Auburn University vehicles used custom designed motor control electronics to process
command information - the Segway RMP vehicle will not require such electronics because
the Segway RMP can be controlled through its built-in CANbus interface. The GUI system
from the Auburn University projects will be modified to gather RMP sensor data and send
RMP control commands over the Segway RMP CANbus. Modifications will incorporate
interface software packages that are provided with the Segway RMP. The modifications will
be done by a team of undergraduate students guided by Dr. Roppel and Dr. Hung.
3. Autonomous Navigation/Sensing: The system hardware will be augmented by the ArcSecond
postion measuring system, with sensors on both the tow vehicle and DGM vehicle. A
dynamic model of the tow vehicle and DGM trailer system will be developed. Information
from the ArcSecond postioning system will be processed by navigation and path following
control software similar to that developed though Auburn University's involvement with the
DARPA Grand Challenge.
4. Implementation :Dr. Hung and Dr. Bevly will supervise the modeling and control system
development. Steering and navigation software developed for the Grand Challenge will be
modified to use ArcSecond data as inputs, and Segway RMP vehicle commands as outputs.
The control algorithm will be redesigned around the dynamic model of the Segway RMP tow
vehicle and DGM trailer system. The GUI program that was developed for manual
teleoperation will be modified to allow switching between manual teleoperation and
1. Eagletron Trackerpod with IO Watchport V camera.
2. One laptop computer to be mounted on the Segway RMP
3. Wireless network PCMCIA card.
4. CANbus PCMCIA card.
5. 802.11b/g compatible radio frequency amplifiers
6. ArcSecond Indoor GPS positioning system
5. Relevant Experience
General robotics activities
Dr. John Hung has been involved with robotics research and teaching activities since 1985, with emphasis
on motion and tracking control using advanced nonlinear methods. He has co-authored nearly 40 papers
related to motion control in refereed journals and international conferences. He is often engaged in
consulting for industry. Applications have included large flexible structures, pointing control, multi-axis
machine control, and motion tracking control. He has regularly taught an undergraduate course in which
students design, implement, and test a line-following miniature robot
DARPA Grand Challenge experience
Dr. Dave Bevly's expertise is in navigation and autonomous control of vehicles, especially with GPS and
inertial measurements. He and his students has been actively involved in the DARPA Grand Challenge,
teamed wth SciAutonics, LLC. His group provided steering and throttle control of the SciAutonics-I
SciAutonics-I/Auburn University vehicle at DARPA Grand Challenge
Reference: R. Behringer, B. Gregory, V. Sundareswaran, R. Addison, R. Elsley, W. Guthmiller, R. Daily,
and D. Bevly, "The Darpa Grand Challenge - Development Of An Autonomous Vehicle," Proceedings of
the 2004 IFAC Symposium on Intelligent Autonomous Vehicles, Lisbon, Portugal.
III. Phase II- Segway RMP Semi-autonomous, Autonomous and Tele-operated application for
Phase I tests in small flat unobstructed areas where path following is the only challenge but the
system acquires laboratory quality repetitive results. Phase II will develop the platform to handle
typical more challenging sites with local obstacles and with areas where the primary positioning
(assumed DGPS for large open areas) is lost.
For Phase II we propose an integrated solution of DGPS with INS, electronic magnetic compass,
and dead reckoning solutions for positioning and a form of synthetic vision for obstacle
recognition and avoidance with a wireless tele-operation and video capability. We envision that
with the on-going technology development and demonstrations being performed to support
robotic activities (specifically the DARPA Grand Challenge), upon completion of Phase I our
proposed technology base may dramatically change. Our first phase II effort is to perform a
technology search and evaluation study to determine the best techniques and equipment to apply,
develop and demonstrate for obstacle recognition and avoidance and to maintain path following
when the primary position is shadowed. For recognition and avoidance, optical recognition, laser
and ultrasonic ranging, and radar are all considerations. Whatever techniques or series of
techniques chosen will require decision algorithms and rules established in the control software to
deviate from the planned path around the obstacle and then to return to the path. When
positioning is lost the system must maintain the pathway until positioning returns and pathway
adjusts can be made. Many techniques are possible to include dead reckoning by using the travel
direction and counting the wheel movements, Inertial Measurement system augmentations,
optical clues, radio, laser, ultrasonic etc. The study will recommend the planned approach for
Phase II . Our proposed cost basis is detailed in the Phase II Technical Details that follow.
The Phase II development approach for semi-autonomous, autonomous and tele-operated
operation will be developed and tested with simulated obstacles and positioning outage at the
McKinley Range Test site.
An active site field demonstration will then be performed at a project of opportunity. The desired
site would be 10-20 acres in size or a typical full week’s field deployment.
1. Results will be documented into a Phase II report, presented at the IPR and the
SERDP/ESTCP Environmental Conference for a Go/No go decision on
enhancement for a Phase III to augment the system to make the system more
Phase II Technical Details
Two new tasks will be the focus of the Phase II effort. First, the system will be modified to
operate over a larger area and under circumstances where a primary positioning system is
temporarily unavailable. Second, a method to detect and avoid above-ground obstacles will be
developed. Many of the technologies and the technical approach will be draw from Auburn
University's past and ongoing experiences with the DARPA Grand Challenge, as well as other
Autonomous Navigation Over Challenging Site In Phase II, the tow vehicle/DGM system will be
modified to navigate over a larger area, and under more general circumstances. Specifically, the
navigation and control system will be based on differential GPS (DGPS) as a primary positioning
system. Additionally, other sensors will be fused through Kalman filters to improve robustness
against short-term loss of the primary positioning system.
1. GPS augmented with Inertial Measurements
2. Dead reckoning using tow vehicle wheel sensors
3. Electronic compass
4. Light detection and ranging (LIDAR)
Technical approach: The navigation and control system will build on Auburn University's Grand
Challenge experience. A GPS based system augmented by inertial measurements through a
Kalman filter has been successfully implemented for the SciAutonics/Auburn vehicle. Additional
sensor data available from the Segway RMP includes left and right wheel displacements. An
electronic compass will also be considered. Simultaneous location and mappping (SLAM) based
on LIDAR data will be examined as a means to maintain short-term navigation and path
following when the primary positioning system is shadowed.
Obstacle Avoidance: The SciAutonics-I vehicle used a combination of LIDAR , ultrasonic, and
optical sensor to handle obstacle detection. The proposed approach for the Segway tow vehicle
will be similar. In addition, the presence of a video camera presents opportunities to integrate
some vision information.
Sensing will consider LIDAR, ultrasound and optical camera
Technical approach: Long-range obstacle detection will be achieved using LIDAR. Short range
detection and location is performed using combination of all sensors. Positions of detected
obstacles will be placed on a 2-D map (bird's-eye view). Path waypoints will be moved as needed
to avoid mapped obstacles.
1. DARPA Grand Challenge: The Grand Challenge vehicle used multiple LIDAR units to
scan both horizontally and vertically. Ultrasonic and optical sensors were also used for
obstacle detection and validation of data.
2. Ultrasound location: Dr. Scotte Hodel and his students have been developing means for
position location by ultrasonic means. That work is currently being applied to 3-D
location of a blimp. Computational techniques for estimating object location are being
3. UAV project: Dr. Hodel examined an obstacle avoidance technique for a project
previously sponsored by USAESCH (low-cost UAV control). That method was more
sophisticated than simply moving waypoint, and is based on optimal control concepts.
Path adjustment is automatically avoided.
1. DGPS: Novatel PROPAK-G2+DB9-424-RT2 with IMU support, 2 cm RTK positioning,
accepts SBAS corrections, real-time DGPS, raw data.
2. IMU: Novatel IMU-G2-H5810 IMU-G2 enclosure with 1 degree per hour IMU for use
3. Waypoint processing software: Novatel SW-PP-GPSIMU Inertial Explorer post-
processing software for GPS/INS applications.
4. LIDAR: Unit made by SICK, used on most of the Grand Challenge vehicles costs about
5. Electronic compass
e) Methodologies. Performance will be bench marked against the speed, productivity and accuracy of man
towed equipment based upon sample runs of the pathways by man towed equipment and by USAESCH’s
extensive experience for productions surveys. Costs and required man power will be directly compared to
current man portable applications.
f) Technical Risks. Because of the phased approach with USAESCH doing the pre-project testing, there is
no risk until Phase II. Phase I is doable with currently available Segway transportation technology and
ArcSecond positioning. The path following routines for a clear unobstructed path following are proven.
The challenge is to refresh and adjust to the desired close tolerance based upon the varying lever arm
between the vehicle and the sensor package. Phase II potentially is very challenging if we were to have to
support all possibilities like for the DARPA Grand Challenge. We plan to chose our solutions that can be
applied to the demonstration sites and to preplan pathway routes to minimize risks. We plan to duplicate
the coverage that could be performed with typical man towed sensor suites such as the Geonics EM-61
g) Related efforts. The Segway® RMP is being developed as a mobile platform for a number of possible
military support missions as part of a DARPA project managed by SPAWAR Systems Center San Diego.
Efforts are being performed by a group of 12 individual Labs and Universities for a wide cross section of
applications as shown on the following web site:
None of the on-going efforts are specific to our proposed project but the findings from several may be
leveraged for the follow on phases where we employ obstacle recognition and avoidance and maintain
position and path following in shadowed areas where we lose primary positioning. In addition the
DARPA Grand Challenge has similar objectives for autonomous vehicle operation but without the added
complexity of path precision and path area coverage and placing a towed vehicle on the desired path.
7. Cost/Benefit of Technology
This project will in Phase I provide a methodology to increase productivity and to acquire more accurate
pathway following that will enhance advanced sensor and algorithm development and enhancement.
Phase II will increase productivity and provide a robotic DGM capability that will enhance safety in UXO
8. Technology Transition
The USAESCH philosophy on innovative technology is to provide for the widespread implementation of
demonstrated/validated technology as soon as proven in the field. This vehicle would be immediately
transitioned to field usage at current and proposed projects where automated DGM can lower cost and
increase safety to personnel.
9. Schedule of Milestones
Pre-proposal (Done) Months -12 to -6
Phase I: Phase II: (Optional)
Develop plans, acquire vehicles Months 1-3 Technology Study Months 10-11
Initial test and modify vehicle Month 4 Technology Development Month 10-18
System Development Months 5-6 Perform demos at a project site Month 18-19
Perform validation testing Month 7 Create Reports Months 20-22
Report Findings Month 8
Go/No Go Decision Month 9 Phase III: TBD
The timeline is based upon a start once funds are received as a 9 month program for Phase I and a 12
month program for Phase II.
U.S. Army Engineering and Support Center (USAESCH), Huntsville, Alabama: Scott Millhouse is the
project lead and principal investigator. USAESCH will be responsible for project direction, project
control, technology oversight, quality assurance. Bob Selfridge, Chief Geophysicist, will design and
implement the sensor testing and demonstration. USAESCH is the Army Center of Expertise for
Ordnance and Explosive and the principal administrator of the OE work to support the DERP FUDS and
Segway and ArcSecond will be equipment and service providers with Auburn University performing
robotic software and sensor equipment integration. The principle Auburn investigator is Dr. John Y.Hung
with support from team members and Associate Professors Dr. David M. Bevly, Dr. A. Scottedward
Hodel and Dr. Thaddeus A. Roppel with support from graduate and undergraduate students.
11. ESTCP Review Comments:
Please elaborate on the proposed uses and concept of operation for this tow vehicle. Discuss
potential terrain and vegetation limitations that could be encountered. The Program Office is concerned
that the applicability for this technology will be limited to only a fraction of the sites where man-portable
systems are required.
This tow vehicle is planned to be used to replace a man for geophysical surveys requiring any man
portable towed sensor to include the Geonics EM series, magnetometer arrays, GEMs, GPR etc. If a
man could safely tow the equipment then we envision that the RMP could perform the task faster,
safer and to a higher accuracy over similar terrain and vegetation coverage. Site pre-planning will be
required. The required pathway must be established by creating a pathway reference file ahead of
time in the field or the office and the file input to the robotic control computer. For Phase II we will
be only monitoring the location of the robot by track maps, video and sensor readings. Future effort
could process, review and interpret the data in real time from a remote office location.
The mobility of the manned Segway exceeded our expectations even with the pavement tires for both
slope and vegetation. The manned proof of concept testing was made in heavy field grass 8-12” high
with no problem. With the ATV tires we envision that it could traverse any location that an EM-61 or
any narrow man towed array could be used for terrain and vegetation. Each drive motor puts out up to
2 HP so there is ample towing power for our limited payload for all reasonable slopes. The Segway
automatically adjusts it’s lean and balances fore and aft but it does not balance or adjust from side to
side. This will necessitate path setup to travel with the slope. If the vehicle travels traverse to a large
enough slope it could upset.
Please reevaluate the cost breakout for Phase II of the proposal. This effort appears to require
significant development work and there is a concern that the planned resources will be insufficient.
Our University partner, Auburn, is a development partner for teams down selected for participation in
the 2005 DARPA Grand Challenge contest demonstration. The costing and equipment has been
reviewed and updated based upon the most recent developments and technology being applied to this
effort. Phase II is planned to demonstrate in typical reasonable conditions to perform a survey by
autonomous, semi-autonomous and by tele-operated control. As in all production surveys, there will
be data gaps that must be filled by manual means such as between trees, boulders and in holes and
Include in your deliverables a draft and final Demonstration Plan, Final Report, and Cost and
Performance Report, in accordance with the full proposal instructions. Guidance describing ESTCP
expectations for these products can be found on the ESTCP web site, www.estcp.org.
Costing is included for labor, travel and supplies to support all ESTCP deliverables to include; a draft
and final Demonstration Plan for Phase I and II, a yearly Phase I Report, a Final Project Report, and a
Final Cost and Performance Report as well as yearly presentation at the IPR, a poster at the
SERDP/ESTCP symposium and required status entry into the SEMS system.
Plan for the first fiscal year of funding to cover work from 1 March through 31 December 2006, and
all subsequent years to cover January 1 through December 31.
Phase I is a 9 month program planned for 1 March through 31 December 2006 with Phase II for 1
January-31 December 2007.
1.2. Cost Estimate
The final composite proposal cost in the requested format is shown in Table 1 with Table 2 showing the
details for our Auburn University partner. USAESCH has contributed approximately $20,000 in pre-
proposal testing and development and will contribute $15,000 in in-house support labor and travel. Phase
I is estimated at $207,333 but with the contribution costs are reduced to $192,333 to ESTCP. It is
independent from Phase II. Following completion of Phase I development and testing a go-no/go decision
is made for Phase II. Phase II is estimated at $292,327 for a total proposal cost to ESTCP of $484,660.