Ambient intelligence in self-organising assembly systems using the chemical reaction model

Abstract

This article discusses self-organising assembly systems (SOAS), a type of assembly systems that (1) participate in their own design by spontaneously organising themselves in response to the arrival of a product order and (2) manage themselves during production. SOAS address the industry's need for agile manufacturing systems to be highly responsive to market dynamics. Manufacturing systems need to be easily and rapidly changeable, but system re-engineering/reconfiguration and especially their (re-)programming are manual, work-intensive and errorprone procedures. With SOAS, we try to facilitate this by giving the systems gradually more self-* capabilities. SOAS eases the work of the SOAS designer and engineer when designing such as system for a specific product, and supports the work of the SOAS operator when supervising the system during production. SOAS represent an application domain of ambient intelligence and humanised computing which is not frequently considered, but therefore none the less important. This article explains how an SOAS produces its own design as the result of a self-organising process following the Chemical Abstract Machine (CHAM) paradigm: industrial robots self-assemble according to specific chemical rules in response to a product order. This paper reports on SOAS in general, the specification of the chemical reactions and their simulation in Maude.

Developing pervasive multi-agent systems with nature-inspired coordination

“…Beside having been conceived as general computational model, chemically-inspired computing can be an effective starting point to realize schemes of dynamic service composition [61] or knowledge aggregation [62]. Network protocols for data distribution and aggregation according to chemical models have also been explored, as in the Fraglets approach [63,64].…”

Modelling and analyzing adaptive self-assembly strategies with Maude

“…The focus of [28], however is not on adaptive systems, but rather on well-formedness scheduling criteria to ensure correctness of statistical model checking techniques (discussed later in Section 7.3). Self-organizing assembly systems, instead, are addressed in [32] using Maude as specification language and the Chemical Abstract Machine (CHAM) [33] as architecture and programming paradigm. CHAM shares some similarities with RRDs in general and our approach in particular, including the use of (reactive) rule-based specifications and nested architectures.…”

“…CHAM shares some similarities with RRDs in general and our approach in particular, including the use of (reactive) rule-based specifications and nested architectures. The approach described in [32], however, does not rely on reflection nor on quantitative analysis.…”

Self-organising assembly systems formally specified in Maude

“…A brief summary is in (Di Marzo Serugendo and Frei 2012). In (Frei et al 2010a) we explained the various relevant concepts, which are briefly reviewed in this article, clarified the relation between Ambient Intelligence and SOAS, and presented the first part of the formal specifications. This work has now been completed and serves as a proof of concept.…”

“…Evolvable and self-organising assembly systems revisit this way of building systems and demonstrate Ambient Intelligence features and responsiveness to people along the following three lines (Frei et al 2010a):…”