Modeling of a hyperheuristic approach within an agent system to support operational planning for industrial product service systems in the production environment

• Applicant: Professor Dr.-Ing. Bernd Kuhlenkötter
• Subject Area: Production Systems, Operations Management, Quality Management and Factory Planning
• Term: from 2019 to 2023
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 424733996

Industrial Product Service Systems (IPS²) offer an opportunity for manufacturing enterprises to offer customized solutions. The basis for the high quality and punctual delivery of IPS² is therefore an efficient short-term (operative) planning. Within the scope of this operative planning, the priority, the execution time and the required qualification and equipment of the service to be planned must be taken into account. In addition, information about the partners in the network is also relevant, in particular about the personnel employed for the provision of services. In addition, information about the personnel involved in the provision of the services is also relevant during planning. The dispatcher is responsible for aggregating this information in the context of operative planning, taking into consideration target parameters such as cost-minimized route planning, high technical utilization and on-time delivery. To support the dispatcher, various software solutions are presently available. However, these do not manage to generate a planning result that takes into account the multitude of relevant information as well as the specific characteristics of the respective IPS² type.The aim of the project is therefore to develop a methodical approach for a decision support system that takes into account the information relevant to the operative planning process, such as the various IPS² types existing in a provider network, and leaves the planner room for personal assessments within the framework of decision making. Since operative performance planning in the context of IPS² is a multi-criteria planning problem that cannot be solved polynomially, a hyperheuristic approach is chosen here. Hyper-heuristics used simpler meta-heuristics and heuristics, so-called low-level heuristics, to solve the planning problem. As a result of this hybridisation, the strengths of individual low-level methods are exploited for specific application areas and their weaknesses are compensated for by other methods. Low-level heuristics suitable for this context will be determined within the framework of the project and the corresponding hyperheuristic will be developed. In this context, the previously selected low-level heuristics are executed by different agents and their deployment is controlled by a superordinate coordinating "hyperheuristic" agent. By using an agent-based approach, the flexibility of the planning tool is increased. Resources and participants of the IPS² network can be mapped by autonomous agents and thus possibly necessary adaptations, such as the integration of further network participants during runtime, can be facilitated. By adapting the boundary conditions of hyper heuristics, the dispatcher also has the opportunity to bring his or her experience into the planning process.

• DFG Programme: Research Grants