Multi-objective stochastic scheduling of job ready times ★ Paul M. Stanfield a , Russell E. King b and Thom J. Hodgson b a ABCO Automation, Inc., 6202 Technology Drive, Browns Summit, NC 27214, USA b Department of Industrial Engineering, North Carolina State University, Raleigh, NC 27695-7906, USA A fundamental scheduling problem is to determine a production start (ready) time based on customer-specified due dates. Typically, the objective is to delay the ready time in an attempt to minimize work-in-process inventory and maximize production system utiliza- tion. In many practical situations, highly variable service times complicate this problem. In such a case, the ready time implies a level of on-time completion confidence for each job. As the ready time increases, the on-time confidence decreases. This paper investigates the ready timejob confidence level tradeoff. A multi-objective model balances the ready time and confidence level maximization goals. The model involves combinatorial and numerical optimization and has an exceptionally complex state space. In view of these complexities, we investigate a pairwise interchange heuristic and a genetic algorithm search solution. Experimental results support solution through a process involving both the heuristic and the genetic algorithm. 1. Introduction and background In most business processes, the sales transaction is conditioned on the ability of the producer to supply the customer by some due date. As a result of this agreement, the producer must determine the time at which raw material must be “ready” and production resources available. For a given job sequence and production system, when the service times are deterministic, the ready time dictates a binary job completion state (either on-time or late). When the production time is stochastic, the ready time implies levels of on-time job completion confidence. The producer desires to maxi- mize on-time completion confidence for customer satisfaction and to maximize the ready time for inventory minimization and production system utilization. The resulting scheduling problem requires simultaneous determination of the production sequence and ready time to balance these conflicting objectives. © J.C. Baltzer AG, Science Publishers ★ This research was sponsored, in part, by Office of Naval Research Grant N00014-90-J-1045 and National Science Foundation Grant DDM-9215432. Annals of Operations Research 70(1997)221 – 239 221