Recirculate tests AI-driven robotic system for EV battery disassembly
The approach uses robotics and machine learning to dismantle batteries from pack-to-cell level. In the first phase, researchers have now developed a set of tools and machine learning models “for the robotised dismantling of an EV battery from pack-to-cell level.” According to the researchers, “it took 18 months of dedicated work by our team of robotics, AI, and ML experts” to complete this first step.
The development is led by Centria University of Applied Sciences in Finland, which has built a robotic cell around a KUKA KR10 industrial robot mounted on a mobile linear track. The robot is equipped with purpose-designed tools and a depth camera to detect, unscrew and remove components of high-voltage packs autonomously.
Multiple machine learning models have been trained to recognise screws, connectors and wiring. Once the lid is removed with a vacuum gripper, robotic tools dismantle internal components while also analysing wire orientations to determine optimal removal strategies.
“There are approximately 50 screws on the lid alone,” explained Tomi Pitkäaho, Principal Lecturer in Research at Centria. “We’ve trained a machine learning model to locate and identify each screw, extract their exact coordinates, and send this data to the robot. With a depth camera installed directly on the tool, the robot can precisely determine not just the x and y position, but also the z-depth for each component.”
The system is not limited to disassembly. A battery identification model has been trained to recognise different battery types even without QR codes or digital product passports. At present, the model can identify packs from Ford and Tesla with near-perfect accuracy, enabling the robot to automatically select the correct disassembly programme.
Recirculate now plans to expand the training dataset to support additional battery types, with the aim of scaling the solution to industrial environments. “This is one of the first working, real-world examples of battery disassembly using machine learning and robotics,” said Pitkäaho. “Until now, most efforts have been purely academic.”
Recirculate was first announced in 2023 and aims to create new business models for the repair, reuse and recycling of second-life batteries. The EU funds the three-year project with 4.9 million euros.
It combines the expertise of eleven companies, including Ford Otosan and DHL, as well as manufacturers of advanced technologies in the fields of AI, intelligent logistics solutions and blockchain technology. Ford Otosan is providing various batteries for the project and is involved in the development of a standardised battery labelling system. DHL is responsible for the safe transport and storage of packs, modules and cells. The other project participants are Probot, Minespider, Eco Stor, Fundacio Eurecat, Libattion, Dafo Vehicle Fire Protection, Iconiq Innovation Limited and Iconiq Innovation Spain and the Swiss CSEM.
0 Comments