Judging Working Rhythm from Body Movement to Prevent Human Errors

摘要

Factories have problems such as accidents and product defects. Accidents makes workers injured, and sometimes takes their lives in the worst case. Product defects cause vendors to lost customer confidence. They also increase production cost because of delays in production processes. The one of main causes for these problems is human errors. The introduction of factory automation (FA) technologies might be one way to solve these problems. However, the cost to introduce FA technologies is so huge that many working processes are still operated by human power. We should provide a method to prevent human error for processes when human power is dominant. Generally, human errors come from lack of worker concentration. This paper values body movement of workers while they are concentrating on works. Workers can record their body movement while working, wearing multiple acceleration sensors on their body. We focus on tasks workers repeat same actions in production lines and so on. When workers engage in these tasks with concentration, they repeat a specific movements because they are getting into a good routine. This paper proposes the working rhythm which show similarity of their current body movement whit those under their concentration. The working rhythm presents the degree of workers concentration. An experiment is conducted to confirm the effectiveness of the proposed method. The experiment makes workers simulated repetitive works tracing a circle shown on a tablet PC display 100 times. We examine the relationship of the working rhythm with the deviation degree of the traced circle. A strong relationship has been found in workers who practice a lot. Workers taking motions similar to those under concentration, draws clean circles. The result has reveals the working rhythm is a good index. etc.