­New article published in 'Computers & Industrial Engineering'

We have recently published our research in 'Computers & Industrial Engineering'. The article is entitled "Scheduling and process planning for the dismantling shop with flexible disassembly mode and recovery level" and deals with the following aspects:

• Novel scheduling problem considers individual disassembly sequences, levels, and destructive modes for multiple product types

• Carbon emissions from end-of-life product recovery and energy usage of the disassembly process are modeled.

• Pareto-optimal solutions for problem with 5 jobs and 4 machines.

• Multi-objective genetic algorithm achieves balanced solutions of makespan and carbon emissions.

• 24 test instances for comparative studies and further research

https://doi.org/10.1016/j.cie.2024.109927

Abstract

The challenges of climate change and resource scarcity have led policymakers around the globe to strengthen the concept of circular economy. Legislation towards extended producer responsibility means that recovery of parts and material from end-of-life (EOL) products is increasingly imposed as a mandatory task for manufacturers across various industries. Consequently, EOL decision-making has emerged as a relevant topic in management science. Recent research has addressed the environmental impact of EOL operations by considering energy usage or carbon emissions from variable disassembly sequences and recovery options. This paper deals with selective disassembly planning and scheduling for multiple products in a dismantling shop and considers the environmental impact that results from the process planning stage. It extends previous formulations by considering flexible disassembly levels and modes. In the multi-objective optimization problem, there is a trade-off between potential time savings, the ecological impact of process-related emissions, and the environmental penalty associated with the recovery of damaged parts, unseparated modules, or unpurified materials. A mathematical formulation is presented and demonstrated using a case study from end-of-life engine disassembly. Next, a multi-objective genetic algorithm (MOGA) is developed and tested using synthetic disassembly problem data. As shown in computational experiments, MOGA outperforms the exact model in more complex problem settings and produces competitive results in a fraction of the time required by the commercial solver.