Simulated annealing approach to solve dual resource constrained job shop scheduling problems: layout impact analysis on solution quality

Real-world manufacturing systems are operating subject to a substantial level of resource constraints. One characteristic model that considers the combination of human and machine resource constraints is called dual resource constrained (DRC). In this context a number of machines nmach is managed by a selection of operators nop, with typically nop ≤ nmach.. A real life case study for an Italian manufacturing company is introduced that uses a set of identical parallel machines being operated by a set of operators. Each job is scheduled to one machine with corresponding loading and unloading process times. A simulated annealing approach is proposed to solve the DRC job shop scheduling problem. A sensitivity analysis is conducted for a selection of algorithm-specific parameters used to solve characteristic DRC layouts. Being characteristic for the just-in-time (JIT) production environment, the high variability in job times has also been taken into account. The results show that the selected layout nmach./nop ratio strongly influences the production system performance. The impact of the ratio of constrained resources has been analysed for different layouts, showing that simulated annealing performs better for single resource constrained problems while also demonstrating that this trend is not symmetrical for different layouts, either operator or machine constrained.