In this master thesis two algorithms for generating a production schedule at a steel producer named Voestalpine Stahl Donawitz GmbH are compared, that could be used as a basis for a decision support system. Three aspects are addressed: the quality of the solution, the computing time and the robustness of the production schedules created. The problem is characterized by a hybrid flexible flow shop, with three constraints, two objective functions to be minimized and stochastic processing times. For the optimization, a deterministic heuristic of Dipak Laha and Subhash Sarin as well as a non deterministic metaheuristic of Dervis Karaboga and Bahriye Basturk, which simulates the behavior of bees during foraging, are used. The aim of the thesis is to find out which of these two approaches is more suitable for solving the underlying problem. In the first part of the thesis the problem is analyzed, mathematically abstracted and defined accordingly to the classification of Vignier. In the second part, the structures of the used algorithms and their modifications are explained in order to solve the present bi-criteria optimization problem with constraints and stochastic data. In the third and last part the concrete implementation of the algorithms in the programming language Java is described and the obtained results are presented and analyzed. The algorithms are tested with actual production data and the two algorithms are compared with respect to the solution quality, the computing time and the robustness of the production schedules created for different problem sizes. The results show that for the stated problem and the given constraints the heuristic of Dipak Laha and Subhash Sarin is not suitable because of its constructive character and the metaheuristic approach of Dervis Karaboga and Bahriye Basturk should be preferred. Finally, an outlook is given on further application possibilities and improvements.