Summary: Genetic algorithms (GAs) have shown to fit the complex needs arising from many tasks of academic optimization for almost 20 years. This paper investigates the application of a GA to a real-world scheduling problem. The objective under consideration is just-in-time completion of jobs. First, the strategic concept of just-in-time production is operationalized according to the applicational demands. Using an extract of real-world dynamic data the proposed GA then serves as a schedule builder. A simulation study shows that the GA clearly outperforms the former way of production scheduling in this application area.
For the entire collection see [Zbl 0842.00026].