ADACOR: A holonic architecture for agile and adaptive manufacturing control

Abstract

In the last decades significant changes in the manufacturing environment have been noticed: moving from a local economy towards a global economy, with markets asking for products with higher quality at lower costs, highly customised and with short life cycle. In these circumstances, the challenge is to develop manufacturing control systems with intelligence capabilities, fast adaptation to the environment changes and more robustness against the occurrence of disturbances. This paper presents an agile and adaptive manufacturing control architecture that addresses the need for the fast reaction to disturbances at the shop floor level, increasing the agility and flexibility of the enterprise, when it works in volatile environments. The proposed architecture introduces an adaptive control that balances dynamically between a more centralised structure and a more decentralised one, allowing combining the global production optimisation with agile reaction to unexpected disturbances.

Introduction

In the last decades world has moved towards a global economy, with markets demanding for products with high quality at low cost, highly customised and with short life cycle, the so called mass customisation. Companies, to remain competitive, need to respond more closely to customer demands, by improving their flexibility and agility while maintaining their productivity and quality, thus imposing significant changes in the manufacturing environment.

Charles Darwin, in his book The Origin of Species, explained that species change over a long period of time, evolving to suit their environment, and that the species that will survive are not the strongest or the most intelligent, but those that are more responsive to change. Translating to the manufacturing world, the companies better prepared to survive would be those better responding to unpredictable and volatile environments, by adapting dynamically their behaviour.

The traditional manufacturing control systems are not designed to exhibit this capability of adaptation and evolution. In fact, their centralised and hierarchical control approaches present good production optimisation, but the rigidity and centralisation of the control structure implies a weak response to change. On the other hand, heterarchical-like manufacturing control approaches present a good response to change and unpredictable disturbances, but as decisions are based in partial knowledge of the system, the global production optimisation is not guaranteed.

In these circumstances, the challenge is to develop manufacturing control systems with autonomy and intelligence capabilities, agile and fast adaptation to the environment changes, prepared to handle efficiently the occurrence of disturbances, and allowing the easy integration of manufacturing resources and legacy systems.

Holonic manufacturing systems (HMS) is a paradigm that translates to the manufacturing world the concepts developed by Arthur Koestler for living organisms and social organisations [1]. Holonic manufacturing is characterised by holarchies of autonomous and cooperative entities, called holons, which represent the entire range of manufacturing entities. A holon, as devised by Koestler, is an identifiable part of a (manufacturing) system that has a unique identity, yet is made up of sub-ordinate parts and in turn is part of a larger whole. The introduction of the holonic paradigm allows a new approach to the manufacturing problem, bringing the advantages of modularity, decentralisation, autonomy and scalability.

In spite of the promising perspective and the research developed by the holonic community, such as referred in [2], [3], [4], [5], [6], [7], and others, compiled in [8], [9], the holonic manufacturing achievements leave some important questions open, as described in [10], namely how to achieve global optimisation in decentralised systems, how should the production control structure evolve to adapt to change, how to specify formally the dynamic behaviour of holonic systems, how to introduce learning and self-organisation capabilities, how to integrate automation resources and how to develop holonic-based control applications.

The proposed control architecture, designated by ADAptive holonic COntrol aRchitecture (ADACOR) for distributed manufacturing systems), intends to contribute to the improvement of the manufacturing control systems performance in terms of the agile reaction to emergence and change, by increasing the agility and flexibility of the enterprise when it works in volatile environments, characterised by the frequent occurrence of disturbances. The focus of ADACOR architecture is the shop floor level and especially flexible manufacturing systems organised in job shop production, characterised by concurrent and asynchronous processes with non-pre-emptive operations and alternative routings. The proposed adaptive architecture intends to be as decentralised as possible and as centralised as necessary, i.e. using a centralised approach when the objective is the optimisation, and a more heterarchical approach in presence of unexpected events and modifications.

ADACOR architecture is based on the holonic manufacturing systems paradigm, and in the following main foundations: decentralised systems, supervisor entities and self-organisation. The manufacturing control functions are in charge of a community of autonomous and cooperative holons, bringing the advantage of modularity, decentralisation, agility, flexibility, robustness and scalability. The introduction of supervisor entities allows the establishment of hierarchies in decentralised systems, to achieve global production optimisation. The introduction of self-organisation capabilities allows the dynamic evolution and re-configuration of the organisational control structure, combining the global production optimisation with the agile reaction to unexpected disturbances.

This paper focuses in the description of the ADACOR control architecture, by indicating the system components, their functions and their interactions, and the adaptive production control model. The formal modelling of the dynamic behaviour of each ADACOR holon class and the synchronisation between the individual models, using high-level Petri nets, is out of scope of this paper, being described in [11].

This paper is organised as follows. Section 2 describes the components of the proposed architecture, referring the holon classes, the supervisor role and the main characteristics associated to each holon. Section 3 describes the interactions among the ADACOR holon classes and Section 4 describes the driving forces of the self-organisation concept, namely the autonomy factor and the propagation mechanisms. Section 5 discusses the adaptive production control approach and Section 6 refers the validation of the ADACOR concepts through their implementation and test in a prototype. At last, Section 7 presents the conclusions.