Decentralized problem solving in logistics and production with partly intelligent agents and comparison with alternative approaches

The paper focuses on the applicability of distributed artificial intelligence (DAI) techniques for solving three “real-world” logistical and production problems. The partly intelligent agents (PIAs) we use are equipped with simple algorithms supplemented by knowledge-based components. By comparison with centralized methods we can judge the benefits of our new DAI-approaches. Since we are working on the solution of real-world problems, the paper is not concerned with ivory tower issues emerging in parts of DAI research. It is a central objective of the work to present in detail how to make “practicable” agent systems and thereby reduce the gap between theory and practice