A new project aims to start tackling the issue with automated robotic disassembly. The nondestructive process, developed at the Fraunhofer Institute for Factory Operation and Automation in Magdeburg, Germany, is designed to provide an alternative to landfills and incinerators, which receive over 80% of generated e-waste.
Instead, the new process involves a robot performing complex actions completely autonomously, taking apart devices including computers to recycle valuable components.
The small percentage of e-waste that currently undergoes treatment is typically shredded, while only a limited portion is manually disassembled, cleaned of hazardous substances, broken down mechanically and sorted into different parts. Those methods are typically expensive and ineffective, however, the Fraunhofer researchers claimed.
“We intend to revolutionise the disassembly of e‑waste. Current solutions require substantial engineering and are limited to a particular product group. In the iDear (Intelligent Disassembly of Electronics for Remanufacturing and Recycling) project, we are pursuing a data-driven methodology so that the widest variety of products, from computers to microwaves to home appliances, can be disassembled,” said Dr José Saenz, manager of the Assistive, Service and Industrial Robots Group at the institute.
At the start of the process, optical sensors and 3D cameras with AI-powered algorithms scan labels with information on the manufacturer and product type. The system detects component types and locations, examines geometries and surfaces, checks screws for rust or other damage and detects any anomalies.
“Previously trained machine-learning algorithms and AI interpret the image data and enable the identification and classification of materials, plastics and components in real time based on sensor and spectral data,” Dr Saenz said.
The data is stored in a ‘digital disassembly twin’, which also provides information on whether a similar product has ever been disassembled.
In the next step, Saenz and his team define the disassembly sequence so the software knows whether to execute a complete disassembly or only focus on the recovery of specific components.
The robot, including robotic arms and a variety of end effectors suited for different tasks, then receives a series of instructions. The disassembly sequences include actions such as screwing, lifting, cutting, extracting, moving levers and cutting wires, all of which can be performed autonomously.
The demonstrator succeeded in removing a motherboard from a computer, which the researchers described as “a very complex task that requires a high level of precision”.
Dr Saenz said: “An AI agent is initially trained to complete the process on the simulation model and later we transfer the trained robot action to the real-world experimental setup. This isn’t necessary for simple skills, such as localisation. We use sensor and camera data for that.”
The team built individual demonstrators for each stage of the process, including a station for the identification and analysis of computers, and AI-generated robot actions to remove motherboards from housings.
They now plan to connect the demonstrators, aiming to integrate all the systems into an automated process.
“Recycling and remanufacturing are key for manufacturing companies to ensure access to raw materials. The recovery of these materials not only reduces the environmental impact of e‑waste but also constitutes a valuable source of raw materials for new products,” Dr Saenz said.
Want the best engineering stories delivered straight to your inbox? The Professional Engineering newsletter gives you vital updates on the most cutting-edge engineering and exciting new job opportunities. To sign up, click here.
Content published by Professional Engineering does not necessarily represent the views of the Institution of Mechanical Engineers.