نظارت بر رفتار کارگران مبتنی بر داده های بزرگ در ساخت متروی چین

Metro construction passes beneath through cities and has characteristics of straitness field and complicated surroundings, which lead to frequent accidents, especially because of worker's unsafe behavior. BBS is one of the construction safety management methods, aiming to observe, analyze and modify worker's behavior. In order to reduce the negative impact of limited behavior observation on BBS, this paper proposes the framework of behavior observation in China metro construction based on big data, which includes: (1) establishing behavioral risk knowledge base; (2) collecting image data reflected worker's unsafe behavior through intelligent video surveillance system and mobile app; (3) establishing big data cloud platform for storing data by distributed file manager. We verified its availability from an example that it can analyze semantic information contained in images effectively, extract worker's unsafe behavior knowledge automatically, retrieve the cloud platform quickly as well. Currently, this research has been utilized in partial construction lines in Wuhan metro and we have got massive valuable behavior observation data.