Intelligent planning for robotic disassembly

Parsa, Soran ORCID: 0000-0002-7182-0001 (2021). Intelligent planning for robotic disassembly. University of Birmingham. Ph.D.

Parsa2021PhD.pdf
Text - Accepted Version
Restricted to Repository staff only until 1 April 2025.
Available under License All rights reserved.
Download (3MB) | Request a copy

Abstract

Environmental concerns and demands, new legislations and rules, and material resource limitations have put pressure on production and manufacturing bodies to seek new strategies and method to meet those criteria. Remanufacturing and reusing of the End-Of-Life (EOL) products has shown a wide potential to deal with production waste and inefficiency. Manual disassembly is not efficient economically and the robotic systems are not reliable in dealing with complex disassembly operations as they have high-level uncertainty. In this research, the geometry of product components and difficulty of disassembly operations were analysed and based on that new optimisation parameters were defined to evaluate the disassemblability of the components. These include Disassembly Handling Index (DHI), Disassembly Operation Index (DOI) and Disassembly Demand Index (DDI). Genetic algorithm optimisation method was modified to find a near-optimal solution. The results using real case study products based on the proposed method shows minimum 10% improvement in manual disassembly time compare with conventional sequence planning methods. In addition, a mathematical model for peg-out-hole disassembly operation in a static equilibrium was developed to determine the relation between peg’s depth, and peg and hole diameter mathematically. Then, force/torque sensor mounter on a TM robot arm was employed to produce force and torque maps for peg-out-hole operation experimentally. The maps were analysed using redundancy method to estimate the peg and hole relative position. The results show the average estimation error using proposed method is 6% which can reliable for peg-out-hole operation. Finally, a disassembly planning method based on human-robot collaboration is proposed. This method employs the flexibility and ability to deal with complex tasks of human, alongside the repeatability and accuracy of the robot. PROMETHEE II method was employed to evaluate disassembly priorities based on the cleanability, repairability, and economy to target the product components based on the remanufacturability criteria. Human-robot collaboration characteristics were defined, and an algorithm was proposed to classify the disassembly task and allocate them to either human or robot. The model encoded using AND/OR method mathematically, and the genetic algorithm was employed to find a near-optimal solution for human-robot collaboration sequence planning. The results compared with Particle Swarm algorithm which shows 8.8% improvement in fitness value. Thus, the developments reported in this thesis have contributed to promoting the efficiency and reliability of the disassembly process as the key step of dealing with EOL products.

Type of Work:

Thesis (Doctorates > Ph.D.)

Award Type:

Doctorates > Ph.D.

Supervisor(s):