Multi-layer edge resource placement optimization for factories

Abstract Introducing distributed computing paradigms to the manufacturing domain increases difficulty of designing and planning an appropriate IT infrastructure. This paper proposes a model solution approach addressing conjoint application resource placement problem in factory context. Instead aiming create exact model, requirements capabilities are simplified, focusing on usability design phase for industrial use cases. Three objective functions implemented: minimizing overall cost, environmental impact, number devices. The implications edge fog considered multi-layer by introducing five levels ranging from on-device, within production system, section, (on-premise), cloud (off-premise). is implemented using open-source modeling language Pyomo. solver SCIP used solve NP-hard integer programming problem. For evaluation optimization implementation benchmark created sample set scenarios varying possible locations, applications, distribution assigned recommendations. resulting execution times demonstrate viability proposed small (100 applications; 100 locations) large (1000 1000 scenarios) instances. A case study section producing electronic components demonstrates practical approach.