AI-Based Robotic Disassembly of E-Waste
The generation of e-waste is predicated to significantly increase, due to the advent of Industry 4.0 and adoption of digital technologies as part of future, sustainable industrial systems.
The use of digital electronic devices is becoming commonplace not only within modern intelligent products but also in the smart manufacturing systems that are used to produce them. Due to design complexity and improved functionality as well as increased demand on quality and reliability, these devices nowadays contain many types of precious metals and other strategically important materials (SIMs). These have led to increased complexity in the task of material reclamation and value recovery from e-waste.
The use of robotic systems as a flexible automation technology, capable of dealing with both part variety and volume, has been investigated. The goal is to produce sorted waste streams with higher concentrations of SIMs through extraction and separation of components and subassemblies, prior to fragmentation and material refining processes. By pre-concentration of SIMs the technical and economic feasibility of materials recovery is greatly improved.
The research methodology is based on three phases of:
- manual non-destructive disassembly of e-waste to define the list of targeted components and subassemblies,
- automated robotic disassembly of components to quantify the throughput rate and waste stream quality achieved, and
- optimisation of robotic disassembly processes based on the design of novel multi-functional tools and flexible fixturing systems.
Professor Shahin Rahimifard - Professor of Sustainable Engineering
Due to their significant economic and environmental benefits, the use of robotic disassembly systems developed in this project will become commonplace in the very near future. Increase in the range of applicability of this approach is being explored through development of AI-based vision systems and multi-functional fixtures and tools to increase in the speed, accuracy, and repeatability of robotic disassembly processes.