To promote environmental sustainability and economic efficiency, the recycling of end-of-life (EoL) vehicles has become a significant industrial activity. The large-scale production of new vehicles necessitates the automation of this recycling process, with robotic solutions emerging as a promising approach. However, this shift introduces new challenges for robotic systems, as disassembly tasks are inherently less predictable and more complex than traditional assembly processes due to the varied conditions of EoL vehicles. This variability places stringent demands on motion planning for industrial robots. This study addresses these challenges by constructing a virtual environment that simulates vehicle disassembly for evaluating motion planning algorithms across diverse scenarios. The research investigates the specific demands placed on motion planners and proposes a set of metrics to assess the quality of the generated trajectories. Additionally, it benchmarks several well-known motion planners using synthetic scenes inspired by real-world use cases, contributing a practical assessment of their performance in this application.
Xiang et al. (Thu,) studied this question.