CONFLICT IDENTIFICATION AND RECONCILIATION IN A COLLABORATIVE MANUFACTURING SCHEDULING TASK

We studied the process of production scheduling in a large chemical plant. Scheduling in that environment is inherently a group process because multiple experts are needed to construct a schedule and to manage its execution. A mathematical formulation of the production scheduling problem yields a mixed-integer linear programming model too large to solve in a reasonable time with current technology. We therefore use an intelligent decision support system (DSS) to heuristically find a satisficing solution to the production scheduling problem. Our DSS is based on a model of the collaborative nature of the task, and it focuses on the communication, argumentation, and reconciliation strategies undertaken by individuals. Using actual production schedules, we show that our DSS can lead to measurable improvements over humanly-designed plans, where the quality of the schedule is measured using the objective function of the mathematical formulation of the problem.