Knowledge-Based Scheduling Systems in Industry and Medicine

PRODUCTION MANAGEMENT IS crucial for achieving the timely and costeffective execution of industrial production processes. In recent years, interest has increased in the use of artificial intelligence technologies for production planning and scheduling.1,2 However, scheduling research typically has been theoretical, has had a narrow focus, and has not covered adaptation to unforeseen events (see the “Scheduling problem” and “Previous scheduling research” sidebars). Our objective has been to use computerbased scheduling systems to enhance the problem-solving capabilities of human domain experts. During our research, we have developed a generic framework for building practical scheduling systems. This framework fosters the reuse of algorithms and the integration of knowledge-based technology into the organizational environment. It also supports dynamic adaptation. We successfully applied our framework in the implementation of three scheduling systems—that is, they all share the same system architecture and use similar problem-solving methodologies. The first two systems deal with serious real-life problems in the manufacturing industry: the rarely investigated continuous-flow scheduling problem and the widely known job-shop problem. The third system shows how concepts and techniques developed for industry can be transferred successfully to a medical domain.