Parametrization of manual work in automotive assembly for wearable force sensing

This research aims to model manual assembly processes by parameterizing operator force readings, specifically for engine and coolant hose connections in an automotive manufacturing line. During automotive assembly, many processes are still performed manually by the human operator due to the complexity of automating the process or product with current technology. Processes include completing hose connections and subsequent “push–pull–push” verification testing. Manual work introduces an opportunity for human error because of natural variation when completing tasks; even a slight inconsistency in operation can lead to an incomplete or missed process. These incomplete processes can pass post-production checks, such as a pressure test, but later loosen and fail, causing rework or warranty issues. To minimize human error, operator force is parameterized to provide real-time feedback to the connection status. The operator force was chosen to classify connections and to verify testing quantitatively. The parametrization was completed by partitioning the shear and normal forces using custom fixtures, with shear being the primary force type required by the process. The varying finger and hand engagement for the different connector locations were factored into the parametrization to encompass a broader range of manually completed tasks. It was found that operator forces in finger engagement for manual assembly could be effectively represented by a limited set of measurable parameters.