Distributed Co-ordination of Steel-making Operations for Reduced Production Stoppages

This paper reports on the development of algorithms for factory decisionmaking and scheduling systems. The particular decision-making and schedule coordination problem studied relates to the synchronisation of a series of events in the context of a production changeover in electric arc steel-making. The paper explores two distributed approaches to this co-ordination problem. The first is a simple manual algorithm that helps co-ordinate the behaviour of steel-plant production via (i) the exchange of relevant inter-unit process information and by (ii) recommending operators take appropriate local actions. The second approach provides a framework for a more general class of problems, and supports either a partially or fully automated co-ordination approach. It is based on an approach from the field of distributed artificial intelligence, referred to as Partial Global Planning (Durfee 1988).