Duc Pham, Chance professor of engineering, Department of Mechanical Engineering, University of Birmingham
After having written about a miscellany of matters ranging from birds, bees, bots and Brexit, to magicians, martial arts, the Muppets and micromanufacturing (yes, once upon a time, I did contribute on the subject of micromanufacturing!), I hope readers of this column will not mind my indulging today in revisiting a technical topic more pertinent to my day job. That topic is remanufacturing or, more specifically, our AutoReman research project, which explores robotic disassembly as a key enabler of autonomous remanufacturing.
Readers may remember that four years ago we embarked on AutoReman with the support of the Engineering and Physical Sciences Research Council (EPSRC). We were initially joined by our industrial collaborators Caterpillar (CAT), Meritor and MG Rover, and research and technology development partners the High-Speed Sustainable Manufacturing Institute (HSSMI) and the Manufacturing Technology Centre (MTC). Subsequently, our team was expanded to include SME partner Recoturbo and academic partners Beihang University and Wuhan University of Technology in China, and the University of Castilla la Mancha in Spain.
It is well known that remanufacturing is a cost-effective means to achieve economic and environmental sustainability. Remanufacturing incurs lower material and energy costs for manufacturers, resulting in lower prices for consumers. Remanufacturing offers better resource security and resilience by reducing the dependence on non-indigenous materials. Remanufacturing also cuts carbon emissions and landfill use, in many cases by more than 90 percent. The long-term goal of AutoReman is to robotise remanufacturing and make it even more attractive to all stakeholders, manufacturers and consumers.
A key step in remanufacturing is disassembly of the returned product. Disassembly is more challenging than assembly to robotise due to variability in the condition of the product: unlike assembly of new or remanufactured products, which is deterministic because the components to be assembled are of known geometries, dimensions and states, disassembly is stochastic as it involves used products of uncertain shapes, sizes and conditions. Thus, disassembly tends to have to be performed manually and laboriously and is a weak link in a remanufacturing system from the point of view of robotisation.
Before AutoReman, there was little fundamental research into robotising disassembly. A common feature of published work on the topic is that it tended to be very much ad hoc and empirical and not supported by a strong scientific basis. To design systems for handling the uncertainties inherent in disassembly, a fundamental understanding of disassembly is required. Such an understanding did not exist and, apart from AutoReman, there has been no attempt to build a knowledge base on disassembly and then apply it systematically to design and build disassembly systems for autonomous remanufacturing.
AutoReman comprises four main work streams. To develop the necessary fundamental understanding, the first work stream focuses on disassembly science and starts with an in-depth investigation of disassembly processes. The second work stream uses the acquired process knowledge methodically to create tools and strategies to enable robotic systems to perform disassembly autonomously. The third work stream concentrates on giving robots the capacity to plan and re-plan disassembly sequences when the initial plan is inoperable because of unexpected problems, allowing them to handle variability and uncertainty.
In many situations, robots can function autonomously alongside people who perform tasks that are too difficult or too costly to automate. In other cases, robots assist human operatives as equal co-workers or as subordinates to humans. The fourth work stream investigates both forms of human-robot collaboration. The assumption is that future disassembly systems will employ cobots, in other words, force-controlled robots that are designed to work safely around people. Work stream four examines how to allocate tasks to such collaborative robots to make the best use of their inherent compliance.
There is additionally a fifth work stream concerned with practical demonstrations. This work stream showcases the individual tools and techniques developed in the other streams, including the strategies for performing elementary disassembly tasks, for planning and re-planning disassembly sequences and for enabling optimal human-robot collaboration. As part of the demonstration activities, a robotic disassembly cell integrating the results of the project is to be constructed to enable the dismantling of complete products. The plan is to demonstrate the disassembly of three selected products remanufactured by our industrial collaborators.
Four years on, what has AutoReman accomplished? We have studied the two operations most common in disassembly, unscrewing and pulling a male component out of a mating female component1, 2. Based on our understanding of those operations, we have devised strategies for robots to perform them. Together with our research partners in China and Spain, we have developed various techniques for automatic disassembly planning, re-planning and human-robot collaboration3, 4, 5, 6, 7. We have built a disassembly demonstration cell with two cobots and used it to disassemble two products, namely a water pump and a turbocharger8.
AutoReman is now on the home straight. However, although all the project objectives should be attained this year, COVID-19 permitting, I believe that there are still many robotic disassembly problems to be resolved. Much work on the science of disassembly still needs to be completed before we can claim to have fully understood it, and many tools and techniques are awaiting development before robots can truly autonomously perform complex disassembly operations. For robotic disassembly research, in Churchill’s famous words: “it is, perhaps, the end of the beginning.”
Acknowledgements I wish to thank the EPSRC, CAT, Meritor, MG Rover, the HSSMI, the MTC and Recoturbo for supporting AutoReman. I am grateful for strong contributions to the project by colleagues at the University of Birmingham, Beihang University, the Wuhan University of Technology and the University of Castilla la Mancha.
www.birmingham.ac.uk/schools/mechanical-engineering/index.aspx
References
1Li, R., Pham, D. T., Huang, J., Tan, Y., Qu, M., Wang, Y., Kerin, M., Jiang, K., Su, S., Ji, C., Liu, Q. and Zhou, Z. (2020). Unfastening of hexagonal headed screws by a collaborative robot. IEEE Transactions on Automation Science and Engineering, volume 17, issue 3, pp.1,455–1,468.
2Zhang, Y., Lu, H., Pham, D. T., Wang, Y., Qu, M., Lim, J. and Su, S. (2019). Peg–hole disassembly using active compliance. Royal Society Open Science, volume 6, issue 8, article no. 190476.
3Laili, Y., Tao, F., Pham, D. T., Wang, Y. and Zhang, L. (2019). Robotic disassembly re-planning using a two-pointer detection strategy and a super-fast bees algorithm. Robotics and Computer-Integrated Manufacturing, volume 59, pp.130–142.
4Liu, Q., Liu, Z., Xu, W., Tang, Q., Zhou, Z. and Pham, D. T. (2019). Human-robot collaboration in disassembly for sustainable manufacturing. International Journal of Production Research, volume 57, issue 12, pp.4,027–4,044.
5Xu, W., Tang, Q., Liu, J., Liu, Z., Zhou, Z. and Pham, D. T. (2020). Disassembly sequence planning using discrete bees algorithm for human-robot collaboration in remanufacturing. Robotics and Computer-Integrated Manufacturing, volume 62, article no. 101860.
6Ramírez, F. J., Aledo, J. A., Gamez, J. A. and Pham, D. T. (2020). Economic modelling of robotic disassembly in end-of-life product recovery for remanufacturing. Computers & Industrial Engineering, volume 142, article no. 106339.
7Fang, Y., Liu, Q., Li, M., Laili, Y. and Pham, D. T. (2019). Evolutionary many-objective optimization for mixed-model disassembly line balancing with multi-robotic workstations. European Journal of Operational Research, volume 276, issue 1, pp.160–174.
8Huang, J., Pham, D. R., Wang, Y., Qu, M., Ji, C., Su, S., Xu, W., Liu, Q. and Zhou, Z. (2020). A case study in human–robot collaboration in the disassembly of press-fitted components. Proceedings of the Institution of Mechanical Engineers, Part B: Journal of Engineering Manufacture, volume 234, pp.654–664.