Deterministic and Stochastic Modelling of a Multi-Type Human–Robot Collaborative Disassembly Line for End-of-Life Electric Vehicle Batteries

Abstract

This thesis addresses the organization of a multi-product human–robot collaborative disassembly line for end-of-life electric vehicle batteries. The objective is to support circular economy strategies by improving the planning of battery disassembly operations, which are essential for reuse, remanufacturing, repurposing and recycling. The studied problem considers several battery types, each with its own tasks, precedence relations, operation times and execution constraints. A deterministic mixed-integer linear programming model is first proposed to determine station opening, task assignment, execution mode selection and resource allocation. Then, a stochastic scenario-based extension is developed to account for the uncertainty of human operation times. The models are tested on instances involving one, two and three battery types. The results show that the deterministic model provides a structured baseline solution, while the stochastic model gives additional information about station-load variability and recourse under uncertainty. The computational study also highlights the limits of exact MILP models for larger instances involving four and five battery types.