This study was conducted at a manufacturing company located in Indonesia that produces Alternating Current Generator Starters (ACGS). In the ACGS production process, there is a solder inspection stage that was previously performed manually by operators. The main issue encountered is incomplete production cycles, a condition where a production cycle is not fully completed, but the product continues to the next stage, increasing the risk of undetected defective or rejected products. Research aims to design a camera check system that integrates the Poka-Yoke method to enhance accuracy and prevent errors caused by human factors based on continuous improvement through the Plan, Do, Check, Action (PDCA) cycle. This study compares two camera inspection programs: program model 1, with individual position calibration per solder point, and program model 2, with a shared calibration setup, to evaluate inspection effectiveness. The quantitative comparative testing results show a detection accuracy of up to 99.92% and an inspection result classification accuracy reached 99.73%, indicating a significant improvement in the reliability of the visual inspection system and quality assurance for soldering results in industry.
                        
                        
                        
                        
                            
                                Copyrights © 2025