| Abstract [eng] |
In this project, the workflow of the manual drive-unit assembly process was examined and a semi automated system was developed to reduce human-error probability and stabilize the final product quality. The main problem was related to the fact that manual assembly operations exhibited high FMEA RPN values, as errors were often not detected in real time, while the assembly of expensive mechanical and electronic components introduced additional risk. During the study, three sequentially operating semi-automated fixtures were designed, forming an integrated assembly system in which Poka–Yoke elements, mechanical positioning features, laser, inductive and photo sensors, step-based Siemens PLC logic, and HMI visual instructions were implemented. Experimental analysis included cycle-time evaluation and the creation of FMEA tables before and after system implementation to determine the extent to which human-factor risks were reduced. It was established that the overall assembly cycle decreased by approximately 16%, and that most manual-process risks were reduced by 62–93%. Depending on the operation, the average RPN value decreased from 224 to 46, representing a reduction of about 79%. The results demonstrated that automated checking mechanisms significantly improved the reliability of screw-tightening, rubber-seal installation, adhesive application, and connection verification. In summary, the semi-automated system enabled a more stable, clearly controlled, and reliable assembly process in which the majority of human-factor related risks were either eliminated or became easily manageable, and the resulting process quality no longer depended on operator experience. The findings can be applied in practical manufacturing environments to reduce defects, increase process stability, and improve the accuracy of quality control. |
