Development of intelligent supervision system for black smoke detection in motor vehicle inspection institutions

To solve the low efficiency of traditional supervision for motor vehicle emission inspection institutions and meet the requirements of the smart inspection and supervision plan of the State Administration for Market Regulation, this research developed a device-edge-cloud collaborative intelligent supervision system for black smoke detection. A customized black smoke dataset for inspection institutions was built, a detection model was trained and optimized based on YOLOv8 and deployed on an industrial edge computing box, and a visual supervision system was developed to connect with Suqian's “Tianyi Cloud Eye” cloud supervision platform. Field tests show the system's recognition rate meets the expected target. It enables full-chain traceability of detection and intelligent early warning of anomalies, effectively improves supervision accuracy, and provides technical references for smart inspection and supervision.