Engineering applications of artificial intelligence
This special issue presents research results on Holonic and multi-agent systems technology applied to manufacturing control, supply chain management and transport network control. These application areas have in common that flows of products (users) have to be coordinated into efficient streams and correct routings. The first paper, by Paul Verstraete et al., titled "Towards robust and efficient planning execution," presents an autonomic schedule execution mechanism. It discusses a Holonic
... acturing execution system that is biased toward following an externally supplied schedule but remains in charge; it discovers alternatives, critically evaluates the given schedule quality and generates information that is absent in the given schedule. It combines the advantage of planning with the robustness of decentralized execution control. The second paper, by Pascal Blanc et al., titled "A Holonic approach for manufacturing execution system design: an industrial application," presents an industrial application. It demonstrates how a Holonic MES copes with real-world complexity and its innovative concepts contribute to industrial objectives. The third paper, by Pavel Vrba et al., titled "Using radio frequency identification in agent-based control systems for industrial applications," discusses a key enabling technology for advanced manufacturing control. The capabilities and limitations of RFID technology have implications for their employment within decentralized control systems, which need dissemination in relevant research communities. The fourth paper, by Joao Sousa et al., titled "Rescheduling and optimization of logistic processes using GA and ACO," scales the control of product flows toward logistic chains. Rescheduling is the key functionality in this special issue in which coping with uncertainties is a central concern. The fifth paper, by Rudy Negenborn et al., titled "Multi-agent model predictive control for transportation networks: serial versus parallel schemes," presents a completely different approach, compared to GA and ACO, in which control theory rather than operations research is the starting point. Note that the control design addresses the need to handle uncertainty and adaptability.