Workforce scheduling algorithms at Grolsch Brewery Enschede

In this thesis, a work force scheduling problem at Grolsch Brewery is examined. A number of tasks must be scheduled over a period of days with a certain number of analysts available to obtain an optimal schedule that results in a maximum number of free analysts available each period. Every task and analyst has its own properties, and there are other side constraints that all have to be taken into account. Therefore, as a �rst do-ability study, an integer linear program is formulated and solved with CPLEX as general purpose integer linear programming solver. This turned out to yield good solutions, yet be very time expensive. Moreover, it turned out that cost considerations prohibit the use of commercial solvers. Therefore alternative, heuristic scheduling algorithms have been implemented for the problem. A simulated annealing approach is used for this. A number of variants have been tested, for example different cooling schedules and di�erent initial temperatures. An extensive computational study of those variations leads to an overview of the results that can be expected to be obtained for the practical deployment for the solution of this speci�c problem.