General functions such as target recognition and intrusion detection are difficult to adapt to the complex railway scenarios, resulting in high false alarm rates. They neither reduce the workload of personnel nor improve accuracy and timeliness. Based on years of understanding and practice in railway construction and inspection operations on business lines, Jiaxun Feihong has developed a railway construction safety intelligent video analysis system. This system is equipped with the Jiaxun Video IoT platform and has developed algorithm models for railway operation scenarios, focusing on the safety of work behavior and strengthening remote online automatic monitoring and analysis of unsafe behaviors of people in key traffic positions and critical posts.

This intelligent work safety analysis system is developed for nine major categories with over 20 scenarios in the construction process of business lines, including workers on duty, violations, crossing lines, early access to the track, out-of-scope work, large machinery operations, foreign objects, and fire/smoke. It models and compares these scenarios to automatically capture and intelligently monitor and analyze violations. This reduces the labor intensity of manual remote online monitoring and analysis, improves monitoring efficiency, ensures worker safety, and forms a safety barrier for railway construction and maintenance operations.