Self-Organising Industrial Assembly Systems
Introduction
Current solutions for industrial
manufacturing assembly systems do not
suit the needs of Mass Customization industry, which is facing low
production volumes, many variants and rapidly changing conditions. An
assembly system is an industrial installation able to receive parts and
join them in a coherent way to form the final product. It consists of a
set of equipment items (modules) such as conveyors, pallets, simple
robotic axes for translation and rotation as well as more sophisticated
industrial robots, grippers, sensors of various types, etc.). An
Evolvable Assembly System (EAS) is an assembly system which can
co-evolve together with the product and the processes (assembly
operations); it can easily undergo small and big changes and seamlessly
integrates new modules independent from their brand or model.This
project extends the notion of EAS with self-organizing principles:
assembly system modules and product parts to be assembled organize
together (among others, choose their coalition partners, their location
and monitor themselves) in order to easily and quickly produce a new or
reconfigured assembly system each time a new product order arrives or
each time a failure or weakness arises in the current assembly system.
Approach
The self-organizing and self-adaptive
architecture that we propose to
use follows a service-oriented architecture. It exploits metadata to
support decision-making and adaptation based on the dynamic enforcement
of explicitly expressed policies. Metadata and policies are themselves
managed by appropriate services. The main point is that the components,
the metadata and the policies are all decoupled from each other and
dynamically updated. Self-organisation and self-adaptation at the
system level is obtained by individual components behaviour being
guided by the prescribed policies. This architecture has been
established in a separate project.
Papers
- G. Di Marzo Serugendo, R. Frei. Experience
Report in developing and applying a method for self-organisation to
agile manufacturing. Technical Report, Birkbeck
College, Dpt of Computer Science and Information Systems, BBKCS-09-06,
June 2009. (pdf)
- R. Frei, G. Di Marzo Serugendo, J. Barata. Designing
Self-Organization for Evolvable Assembly Systems. Technical
Report, Birkbeck College, Dpt of Computer Science and Information
Systems, BBKCS-09-04, June 2009. (pdf)
- R. Frei, G. Di Marzo Serugendo, J. Barata. An Architecture for Self-Managing Evolvable
Assembly Systems. IEEE
International Conference on Systems, Man and Cybernetics (SMC'09), San
Antonio, Texas, October 2009 (pdf)
- R. Frei, G. Di Marzo Serugendo, J. Barata. Designing
Self-Organization for Evolvable Assembly Systems. Second IEEE
International Conference on Self-Adaptive and Self-Organizing Systems.
Isola di San Servolo (Venice), Italy, pp. 97-106, October 20-24, 2008. (pdf)
- R.
Frei, J. Barata, G. Di Marzo Serugendo. A Complexity Theory
Approach to Evolvable Production Systems, In Proceedings of the
International Workshop on
Multi-Agent Robotic Systems (MARS 2007), Peter Sapaty and Joaquim
Filipe (Eds),
INSTICC Press, pp 44-53, May 2007, Portugal (pdf)
Poster
- Self-Organisation in Evolvable Assembly Systems. 50th Years of
Computing at Birkbeck. (pdf)
Grants
Partners
- Uninova, Portugal
- Birkbeck College
G. Di Marzo Serugendo
Sep 2009.