This MDPI Robotics article introduces a hybrid task planner that coordinates multiple robotic arms to dismantle electric vehicle battery packs efficiently. The planner fuses logical reasoning with motion planning so that robots can share a workspace without collisions.
Planner Design
- End-effectors capture RGB data that feeds a perception module for locating fasteners and modules.
- Logical task sequencing assigns actions to robots while considering grasp availability, reachability, and safety rules.
- Motion planners (RRT, RRTConnect, RRT*) execute the sequences; results showed near-identical completion times across planners, confirming the task decomposition is the main efficiency driver.
Key Outcomes
- Demonstrated on an EV battery pack with significant structural complexity.
- Achieved ~540 s disassembly time across planners, validating robustness to planner selection.
- Highlights how multi-robot coordination can generalise to other disassembly scenarios in the circular economy.