A heuristic-based car shop scheduling application

The formulation of a heuristic-based car-shop scheduling application is described. The problem involves scheduling repair jobs on cars, given restrictions on operator availability and other resource/time constraints. The problem is solved by taking an intelligent generate-and-test approach and extending the simple notion of scheduling-the allocation of resources to tasks over time within constraints defining the system. Dispatch of tasks for scheduling and allocation of resources to them are guided by a set of heuristics. The system is built on a user-extensible knowledge base of rules and heuristics written in Prolog. The emphasis in the system is on providing a flexible AI problem representation and also collecting some empirical results on the performance of different heuristics in the system