Today, manual dismantling can take more than one hour. RESTORE aims to support automation approaches that could significantly reduce this time, while improving safety and consistency.
Researchers at UIA, building on resources and knowledge from RHINOCEROS, are testing new vision models allowing them to achieve over 95% accuracy in identifying large battery components. Further refinements are underway for smaller components (e.g. screws).
Additional to this advanced component identification, UIA has established the operation sequence and task recovery protocols for autonomous robotic dismantling. With off-the shelf tools not always suitable for dismantling operations, UIA researchers have also been developing 4 custom disassembly tools – already functional.
Other work packages have started with basic manual trials for unwinding and delamination on small batches of prismatic cells – work which will feed practical information for later automation dismantling processes. First observations provided by KYBURZ already flag the presence of adhesives and plastic tapes, which could later request adapted strategies and tools for the autonomous dismantling.
© visual: Adobe Stock Photos
