A Monitoring and Control System Based on Image Recognition for Iron Filings Cleaning of CNC Machine Tools

This study addresses a key challenge in the CNC tool machining industry: automated in-machine chip removal. Initially, we designed and manufactured an innovative nozzle, incorporating elastic plastic hoses to replace sections of traditional hard coil pipes. A 3D-printed auxiliary mechanism was developed to enable the servo motor to direct the nozzle to any specified coordinates on a two-dimensional plane. Simultaneously, captured images are divided into multiple subblocks for detailed analysis. Image recognition algorithms process each block to determine the location of iron filings. The ratio of detected iron filings is then computed to establish cleaning priorities, which the embedded system uses to direct the servo motor for effective cleaning. Additionally, the system features a human-machine interface that monitors the cleaning process, providing on-site personnel with real-time insights into the machine’s internal conditions. Experimental results from a custom simulation environment demonstrate that the system efficiently reduces the accumulation of iron filings on the platform to levels below the set threshold within a short time frame. This research aims to minimize machine tool idle time, lower personnel costs, and enhance overall production capacity.