Document Sample
Assembly Powered By Docstoc
					                  European Assembly:
                Opportunities or Threats?

                                   Mauro ONORI
                              Luis Camarinha-MATOS
                                   Jose BARATA

                                 ISATP 2003
                         Besançon, France, July 11th, 2003
M.Onori, KTH July 2003
   Assembly Net is a Thematic Network which aims
   to broker knowledge, solutions, and requirements
   to and from its members.

   Its focus is on mini & micro-assembly, although
   the transitions between macro-micro, and its
   ensuing problems, are also covered.


M.Onori, KTH July 2003
  In order to tackle the challenges
  being posed by the current
  trends and future prospects, the
  Assembly Net consortium has
  created its own R&D roadmap
  for the next decade.

  Some results & conclusions from
  this endeavour will hereby be

M.Onori, KTH, July 2003
                                       Background facts

  Is there anything to worry about?

     20% of manual assembly has left Europe since 2000.

     Production Engineering schools are losing students.

     The birth rate in Europe has been negative.

  In the meanwhile….

     Samsung has a ”private” university with 30,000 students.

     China produces 4,000,000 engineers/year...

M.Onori, KTH, July 2003

  Products usually consist of macro, mini, and micro parts,
  but what are the general subdivisions?

M.Onori, KTH, July 2003
                          Knowledge Status

M.Onori, KTH, July 2003
                          Knowledge Status

M.Onori, KTH, July 2003
                          Knowledge Status

M.Onori, KTH, July 2003
                                            The Objectives

       The Roadmap was drawn out in order to:

       • Give an idea to our members as to what the
         future (required) R&D areas will be.

       • Confirm, and pinpoint, that there are gaps in the
        current European R&D portfolio and industrial

       • Analyse whether certain trends pose serious
         threats or may offer opportunities.

       • Offer a Vision and details for attaining the needs.
M.Onori, KTH, July 2003
                                                        Trend 1

Lack/Loss of Process Knowledge
       Assembly, as a process, is not structured, not classified,
       and encompasses a huge range of sub-processes that
       have not been categorised.

       This affects standardisation, modelling, and modularisation
       efforts, and supportive measures such as ontologies
       cannot be verified.

       Since academic research is becoming less involved in
       application-oriented projects, and manual workers are
       being laid off in industry, this lack of process knowledge
       is being deteriorated (further loss).

M.Onori, KTH, July 2003
                                             Trend 1

   In order to standardise equipment, one needs to
   formalise the process it is designed for.

   Formalising a process requires in-depth knowledge
   of it, and all its sub-processes.

   Therefore, it is fairly safe to state that no truly
   modular or standardised assembly equipment can
   yet be achieved.

   Note that ”assembly” encompasses anything from
   spot-welding to micro laser joining techniques, often
   including testing, packaging and marking.
M.Onori, KTH, July 2003
                                                      Trend 2


      Outsourcing, a strategy used to curb costs and focus on
      what companies call “core competencies” …

      The average wage of a south-east Asian worker is of
      €100-150 / month.
      We simply cannot develop a hardware solution that can
      compete with these figures.

      In the meantime, there is great confusion within
      companies as to what IS their ”core competence”…

M.Onori, KTH, July 2003
                                                   Trend 2

      In a questionnaire sent out by the Assembly-Net, most
      Members replied that assembly WAS a core competence…

M.Onori, KTH, July 2003
                                                       Trend 2

          … they also stated that assembly, including some
          final assembly, was being outsourced…!

M.Onori, KTH, July 2003
                                                 Trend 2

     The leading Contract Manufacturers are not European,
     which entails an export of process & product knowledge.

M.Onori, KTH, July 2003
                                                         Trend 3

Weak link between Product-Production System Design
  DFA, DFA2, DFMA, and other tools have not
  had the intended depth of impact.

  Products are still, in general, not being designed with particular
  production system requirements in mind, or vice-versa.

  Customer requirements are fed back too late, production
  system requirements are not well understood, etc.

  For products with parts and proceses which defy the
  human eye, and open new engineering domains, this will be

  There is no DFμA…
M.Onori, KTH, July 2003
                                                        Trend 3

  In fact, product design is starting to be absorbed by the Contract
  Manufacturers as well.

M.Onori, KTH, July 2003
                                                        Trend 4

Very Low Equipment Re-Usability

   Unfortunately, the existing paradigm of highly flexible
   assembly systems still prevails and results in expensive,
   highly technological solutions, even when tagged with the
   feature “modular”.

   Resulting fairly adequate to many different product types,
   they fail to be very performant in any domain.

   What is really needed are systems that may evolve with the
   product, market, and technological changes that a company
   goes through…

M.Onori, KTH, July 2003
                                                       Trend 4

     Cost was given as one of the main reasons for competitivity…

M.Onori, KTH, July 2003
                                                      Trend 4

 Re-configurable systems are a popular issue today.

 Modular, Plug & Produce systems can be achieved, but what
 are also required are new investment approaches, cost models,
 methodologies, and supportive services.

 Once again, modular systems can only be achieved on the basis
 of the processes they serve. Without sound process knowledge
 only mechanical modularity may be achieved.

M.Onori, KTH, July 2003
                                                  Trend 4

Modular cell components must be linked to the
assembly processes that need to be carried out.

Standards need to be developed.

Ontologies need to be drawn out.

M.Onori, KTH, July 2003
                                                            Trend 5

 Maintenance equation
        Maintenance issues have not been regarded as important
        by the R&D community.

        This is serious since operational information is of capital
        importance and should absolutely be fed back into the
        other product lifecycle phases.

        At present, the main knowledge in maintenance is held
        by shop-floor operators and foremen, people that have
        been losing their jobs at an alarming rate.

        Note that product and assembly system re-engineering
        and re-configurability require serious maintenance data
M.Onori, KTH, July 2003
                                                           Trend 5

   The advent of micro and nano-factories will obviously worsen
   this situation, mainly due to their extremely small size: human
   intervention in case of production failures will be impossible.

   The risk that entire microfactories will be simply disposed
   of in case of fatal production failures is large. This, in time,
   will become an eco-sustainability issue.

   This will be a multi-disciplinary problem…

M.Onori, KTH, July 2003
                                                            Trend 6

 Product Lifecycle uncertainties
The symptoms manifest themselves in very low assembly system
adaptability to product and production variations, the inability to
forecast product changes in time, etc.

The somewhat hidden problem, however, is that the major part of
producing companies have to deal with planned products and
existing production facilities.

Ideally, they would like to fit any new product, or product variant,
into an existing assembly system with as low costs as possible.

This entails that any new assembly system solution has to fit into
an existing facility, even though this facility may denote
performance limitations. The same applies for new product designs.

M.Onori, KTH, July 2003
                                                      Trend 6

Another major problem is that current control/supervision
architectures are not efficient: any change or addition made at
shop-floor level requires programming modifications (production
This implies that qualified programmers must be called in since
this competence is seldom found within producing companies, let
alone SMEs. To worsen the situation, such program modifications
(even small changes) might affect the global system architecture,
therefore causing an increased programming effort and the
potential for unwanted side-effects.

In order to reverse these negative trends,a new methodology
which encompasses all phases in a product lifecycle is required.
This goes beyond linking product design and system design.

M.Onori, KTH, July 2003
                                                        Trend 7

 Virtual Enterprising / Collaborative Networks
   Multi-agent technologies and architectures are becoming
   a pre-requisite for enterprise agility.

   These solutions may be adapted to assist the development
   of re-engineering architectures, shop-floor control systems,
   and other aspects.

   However, this is being blocked by:

   • The lack of assembly process knowledge (cannot be modelled).
   • Difficulty in integrating different enterprise cultures.
   • Difficulty in bridging over to the Production Engineering

M.Onori, KTH, July 2003
                                                        Trend 8

   Social & Educational Issues

      Demographically speaking, Europe will face a very
      serious decline. There will be less workers and less
      students, all the while other world areas are increasing
      in these sectors.

      Mini and micro assembly also requires a re-thinking of
      traditional Production Engineering practices, in which
      Computer science, Human factors, Economics, and Natural
      Sciences will have to be absorbed.

M.Onori, KTH, July 2003
                          Trend 8: survey

M.Onori, KTH, July 2003
                          Trend 8: survey

M.Onori, KTH, July 2003
                                                Trend 8: survey

Preliminary Conclusions (survey to be finalised September 2003):
The USA is conducting R&D work in areas which are more
appropriate for the enhancement of micro assembly than Europe:
Positioning, handling, microfactory, etc. 1-2 large scale efforts.

Europe is still very diversified and too focussed on particular
problems, thus missing the overall picture. No large-scale efforts.

No R&D work seems to be active within the services aspects and
process structuring aspects (vital for re-configurable assembly).

Europe does not embedd IT and other disciplines well enough into
its Production Engineering courses.

No work being done on ontologies for micro assembly.

M.Onori, KTH, July 2003
                                   Other underlying threats

           •   Outsourcing <> enterprise networks.
           •   Outsourcing = Problem shifting.
           •   Micro & nano= unstable processes.
           •   Micro & nano=low visibility.
           •   No process structures = no standards.
           •   Poor knowledge management.
           •   Demography.
           •   Miniaturisation<>outsourcing.
           •   Eco-sustainability.
           •   New materials=disassembly?

M.Onori, KTH, July 2003
                                       GAP 1

 There is no concerted R&D effort to structure,
 stabilise, formalise and categorise assembly

 This is vital for the standardisation work,
 applications requiring process models, and

M.Onori, KTH, July 2003
                                       GAP 2

  Maintenance and product design require new,
  dedicated tools for the micro-assembly

  Micro processes are inherently invisible and
  unstable, which means that Gap 1 must be
  covered before, or simultaneously, with Gap2.

M.Onori, KTH, July 2003
                                         GAP 3

     The trends detailed, as well as Gaps 1&2
     clearly point out that there must be an
     effort to develop an integrative

     This requires the development of ontologies
     and system & control architectures.

M.Onori, KTH, July 2003
                                        GAP 4

   The academic infrastructure for Production
   Engineering in Europe must stop adapting
   traditional approaches to the micro domain.

   New, far more multi-disciplinary courses
   and R&D efforts must be established.

   e.g- μcontrol issues require IT, economics,
   contract law, and social behavioural models…

M.Onori, KTH, July 2003
                          The Vision

M.Onori, KTH, July 2003
                                             The Vision

     The Evolvable Ultra Precision Assembly Systems
     vision is based on a methodology that enables
     the continuous evolution of an existing assembly

     It may evolve from manual to automatic or
     vice-versa, depending on the needs.

     It will assist product designers and system

     It will exploit operational data and after sales
M.Onori, KTH, July 2003
                                            The Vision

 The basis for exploitation of this methodology
 resides within virtual engineering architectures.

 The formation of adequate knowledge management
 foundations will rely on the creation of virtual R&D
 centres that could link tools and information (resource
 providers), to the users (producing companies)
 through coordinated sets of decision-making
 mechanisms and data validation tools (contracts).

 Such an architecture can then be applied to the
 re-engineering problem:
 a shop-floor controller for agile re-configurability.

M.Onori, KTH, July 2003
                                        Product 1

        The standard set of                             add/remove
        task-oriented components.                       standard components


                                                          Products 1 & 2

                                                        standard components

     A methodology that supports each
     phase of the product lifecycle.

                                                    Products 1, 2 & 3
M.Onori, KTH, July 2003
                                        New Paradigm

       This represents a shift in thinking since
       it implies that theoretically very flexible,
       multi-purpose cells will be replaced by a
       highly flexible concept consisting of
       several well-targeted but not, in
       themselves, highly flexible components.

M.Onori, KTH, July 2003
                                                    New Investment Approach

 Precision Assembly
     Operations                     Hardware Depot

                                                          Hardware Depot
                  Hardware Depot

                                   Hardware Depot

M.Onori, KTH, July 2003

    This approach is being materialised within the
    EUPASS (Evolvable Ultra Precision Assembly)
    Integrated Project within the EC’s 6th Framework.

    Other initiatives are also taking form, from national
    projects to larger consortia.

M.Onori, KTH, July 2003

 In view of the rate at which assembly is disappearing
 to low-wage nations, students desert Production
 Engineering disciplines, and the lack of assembly
 process structuring, EUPASS represents a vital yet
 small European initiative.

 European academia should focus to a far greater
 extent to the coming technological requirements,
 and adapt its intellectual infrastructures to meet these

M.Onori, KTH, July 2003

Shared By:
vivien renata vivien renata