Knowledge-based scheduling of duty rosters for physicians.

Applications of artificial intelligence methods to problems of common sense in medicine are rare. Our approach deals with a special class of resource allocation problems concerning fairness in the everyday life in a hospital. We treated the construction of a duty roster in a medical environment. We developed a fairness reasoning machine embedded in the expert system PEP for constructing a duty roster. Furthermore, we elicited and generalized knowledge about fairness between physicians. PEP has been used routinely. We observed a clear short cut of work time of the user in running it, and a 'fair' long-term allocation of physicians in the duty roster.