4 Presentations

• 
  10:30 AM - 10:52 AM

  A Hybrid Approach to Large-Scale Short-Term Scheduling in Make-and-Pack Production

  • Philipp Baumann, presenter, University of Bern
  • Norbert Trautmann, University of Bern

  We investigate short-term scheduling of industrial make-and-pack production processes. The planning problem consists of minimizing the production makespan while meeting given end-product demands. A large number of operations, sequence-dependent changeover times, multi-purpose storage units with finite capacities, batch splitting, quarantine times, partial equipment connectivity, and material transfer times render the problem a challenging task. Known MILP formulations for such production processes can solve only small and medium-sized problem instances in reasonable CPU times. In this paper, we present a hybrid heuristic approach to tackle large-scale instances. Under this approach, the set of batches is divided into several subsets according to a priority rule. The subsets are then scheduled iteratively using a MILP model. We enhance the performance of the heuristic by eliminating redundant constraints and variables after each iteration. The applicability of the proposed heuristic is demonstrated by means of a real-world production process.

• 
  10:52 AM - 11:14 AM

  Perspectives on Interactive Metaheuristics

  • David Meignan, presenter, University of Osnabrück
  • Jean-Marc Frayret, Polytechnique Montréal, CIRRELT
  • Sigrid Knust, University of Osnabrück

  This paper investigates interactive metaheuristics as a way to support the integration of metaheuristics in real applications. The study focuses on the use of metaheuristics in optimization-based decision support systems. In spite of the fact that metaheuristics have been proven to be successful as real problem solvers in a variety of application areas, several limitations restrain the use of metaheuristics in such decision support systems. These obstacles to a practical use of metaheuristics may be overcome by taking advantage of human expertise during the solving process. In this context, this paper investigates interactive metaheuristics. In this original approach of metaheuristics, the interaction with the user aims at guiding the search process, enriching the problem model, involving and informing the user on the decisions taken by the optimization process. In this paper, we first try to identify the main obstacles to a practical use of metaheuristics in optimization-based decision support systems. Then, we present an interactive approach of metaheuristics and propose a conceptual framework to analyze existing interactive approaches.

• 
  11:14 AM - 11:36 AM

  Interactive Planning System for Forest Road Construction

  • Jean-Marc Frayret, presenter, Polytechnique Montréal, CIRRELT
  • David Meignan, University of Osnabrück
  • Gilles Pesant, Polytechnique Montréal

  In this paper, we propose an interactive decision support system for forest road construction planning. This approach is based on an interactive heuristic approach, within which the user contributes in a cooperative manner to the optimization process. The objective is to exploit the problem-domain expertise of the user in order to generate more realistic solutions that integrate aspects not captured by the heuristic objective function. This paper presents the user interface, the interaction mechanisms and the heuristic developed to support the cooperation between the computer and the user. We present experimental results based on real problem instances. A comparison shows clear advantages for using the proposed interactive approach over a pure manual approach.

• 
  11:36 AM - 11:58 AM

  A Three-step Approach for Operating Room Planning and Sequencing under Uncertainty

  • Aida Jebali, presenter, Ecole Nationale d'Ingénieurs de Tunis
  • Sana Aniba, Ecole Nationale d'Ingénieurs de Tunis

  A stochastic three-step approach is proposed for operating room planning and sequencing. Two types of uncertainties related to operating room activities are taken into account: (1) uncertainty related to surgery case duration and (2) uncertainty related to the daily arrival of emergency surgery cases. After constructing operating room plan and fixing off-line and on-line sequencing rules, simulation is used to evaluate the performance of the operating rooms. The obtained results highlight the importance of using a stochastic approach for operating room planning. Indeed, regardless the used sequencing/resequencing rules, operating room schedules obtained by the stochastic approach are more robust than the ones obtained by a deterministic approach. Besides, the utilization cost of operating rooms is reduced when a stochastic approach is used.