Automated IE System to Observe Worker´s Behavior with Low-Cost Sensors

A serious shortage of IE specialists and the high cost of IE implementations are spurring demand for automated IE systems. The conventional automatic method, using knowledge concerning the worker\´s role and the factory layout, recognizes the "basic operations" which are observed by IE specialists in traditional IE. However, the method, in many cases, requires the IE specialists to discover the causes of the inefficiency. We propose a new method to identify the causes automatically. We define the "basic situations," which include the causes. The fundamental idea of the method is that the system translates the basic operations to basic situations by utilizing peripheral information such as product state and other worker\´s state. Our experiments in an actual factory showed the proposed method effectively identifies the causes of the inefficiency in the factory