Fostering End-of-Life Utilization by Information-driven Robotic Disassembly

Abstract Economic and ecological feasibility are the key factors for companies to engage in circular economy processes. Disassembly plays a major role in this environment, as it is one of the most complex and labor-intensive steps in end-of-life processing. For this reason, an automation of the disassembly in order to facilitate performing the necessary tasks is highly favorable. In this contribution, an agent-based robotic disassembly system, evolved from the innovation cluster “Recycling 4.0”, is presented. The proposed system features a novel information-driven control architecture, combining a superordinate, cloud-based knowledge base and comprehensive sensory perception. In pursuance to find an optimized utilization strategy for each part to be disassembled, various features, such as individual life-cycle data, material composition and optical rating are taken into account for an artificial intelligence based multi-criteria assessment in order to determine an appropriate end-of-life option for each of the part’s components. It is shown that the developed system is able to operate in a collaborative scenario of an electric vehicle traction-battery disassembly, deciding upon the level of disassembly for the end-of-life option chosen based on the available information, which is acquired through a system-wide interoperability standard. Being able to actively contribute to the overall knowledge base of the framework by processing validated product- and process-knowledge back to the knowledge base, the robotic system fosters a feasible progression towards closed-loop supply-chains in a circular economy and enables new market participants to engage in recycling operations.