Designing collaborative robot workstations for human-centred automation in final assembly - A task allocation approach

Doctoral thesis, 2023

With the rise of increasingly intricate products, human operators are stretched thin by greater physical and cognitive demands. A growing necessity has also arisen for automation to assist these operators supportively and productively while maintaining optimal efficiency. One solution to this problem is the integration of collaborative robots, specifically, using task allocation to apply collaborative robots in the design process correctly. This thesis uses a task allocation approach to design collaborative robot workstations.

This thesis addresses two questions: 1. how humans and robots collaborate in manufacturing and final assembly operations, and 2. how human-centred task allocation supports the design of collaborative robot workstations in final assembly. The first part addresses how collaborative robots are being used in manufacturing and assembly operations and what needs to be done to improve their adaptation in a way that increases human-centred automation where human operators are actively working with the robots. The second part addresses the design process of human-robot collaborative workstations using levels of automation and task allocation approach.

Findings reveal that there is great interest from industry to use collaborative robots. Some companies use them today, but mainly as a low-cost automation solution. The active collaboration between humans and robots is almost negligible. To achieve the full potential of collaborative robots, active collaboration needs to be improved, which is mainly lacking due to a lack of clarity on achieving human-robot collaboration. With the use of task location and levels of automation, the issue of the need for more clarity can be solved. This presents a new level of automation matrix specifically developed for collaborative robot applications that use the physical and cognitive skills of both humans and robots. This new matrix is exemplified using a wire harness assembly process, where cognitive and physical levels of automation are allocated to different tasks during hierarchical task analysis.

The industry is showing a great interest in implementing collaborative robot applications, and this thesis shows it. As the development of collaborative robots continues, research efforts are generating new concepts for their integration. It is evident that robots working alongside humans have the potential to revolutionize final assembly. Nonetheless, for a successful integration, it is vital to consider both human and technological factors. This thesis serves as a starting point towards that direction.