No abstract

No abstract

An efficient production planning and control is inevitable for the economic operation of a manufacturing system. As an essential part of production planning and control, process scheduling aims to assign processes to the available resources of a manufacturing system under consideration of its objectives. This forms an optimization problem also known as job shop scheduling which can be solved with computer aided techniques. Exact solution methods are only practical up to a certain number of functional units and processes, therefore, approximation methods are used in industry. However, as the problem size increases, the computational solving time increases significantly and the solution quality decreases in equal measure. In order to react to the current effects of global crises, such as disruptions in supply chains, approaches for fast and efficient rescheduling are needed. A dynamic shop job scheduling approach using Quantum Annealing bears the potential to close this research gap. Previous work has shown that Quantum Annealing is able to solve static job shop scheduling problems within seconds while finding good solutions. However, in a flexible environment such as a manufacturing system, the static approach is not suitable for process scheduling. Therefore, a dynamic Quantum Annealing-based approach for job shop scheduling with consideration of machine breakdowns and new job arrivals is proposed. The approach monitors a manufacturing system and reacts to changes in the job pool or availabilities in functional units with rescheduling. The method is tested with several use cases involving small and large-scale problems and is compared with a simulated annealing approach. Thereby, the Quantum Annealing-based computations show better results regarding solution quality and computing time. Moreover, the dynamic approach bears the potential for industrial application, especially as a supplement to a conventional advanced planning system.

Das auf quantenmechanischen Prozessen basierende Quanten-Annealing ist eine Technologie, die es erlaubt, Energieminimierungsprobleme effizient zu lösen. Durch die Formulierung von Reihenfolgeplanungsproblemen als Energieminimierungsprobleme bieten sich Potenziale einer zeiteffizienten Lösung mittels Quanten-Annealing. Gegenstand dieses Beitrags ist ein Konzept zur Überführung von Reihenfolgeminimierungsproblemen in eine mittels Quanten-Annealing verarbeitbare Problemformulierung.

Factory layout planning aims at finding an optimized layout configuration under consideration of varying influences such as the material flow characteristics. Manual layout planning can be characterized as a complex decision-making process due to a large number of possible placement options. Automated planning approaches aim at reducing the manual planning effort by generating optimized layout variants in the early stages of layout planning. Recent developments have introduced Reinforcement Learning (RL) based planning approaches that allow to optimize a layout under consideration of a single optimization criterion. However, within layout planning, multiple partially conflicting planning objectives have to be considered. Such multiple objectives are not considered by existing RL-based approaches. This paper addresses this research gap by presenting a novel RL-based layout planning approach that allows consideration of multiple objectives for optimization. Furthermore, existing RL-based planning approaches only consider analytically formulated objectives such as the transportation distance. Consequently, dynamic influences in the material flow are neglected which can result in higher operational costs of the future factory. To address this issue, a discrete event simulation module is developed that allows simulating manufacturing and material flow processes simultaneously for any layout configuration generated by the RL approach. Consequently, the presented approach considers material flow simulation results for multi-objective optimization. In order to investigate the capabilities of RL-based factory layout planning, different RL architectures are compared based on a simplified application scenario. In terms of optimization objectives, the throughput time, media supply, and clarity of the material flow are considered. The best performing architecture is then applied to an industrial planning scenario with 43 functional units to illustrate the approach. Furthermore, the performance of the RL approach is compared to the manually planned layout and to the results generated by a combined version of the genetic algorithm and tabu search. The results indicate that the RL approach is capable of improving the manually planned layout significantly. Furthermore, it reaches comparable results for the throughput time and better results for the clarity of the material flow compared to the combined version of a genetic algorithm and tabu search.