Acquiring After-Sales Knowledge from Human Motions

Abstract

Human motion is a key property to understand human’s knowledge and intention. A service technician moves to achieve his duty: facility maintenance. The series of his motions is the output of his observation, diagnosis, and repair. Therefore the motions are recognized as a knowledge source. From the viewpoint of after-sales productivity, improving service technicians’ skill is a vital key to quick and efficient service. However recording service cases costs high and is a bottleneck of knowledge and experience sharing systems. We propose a framework that measures human motion behaviors and elicits knowledge from the motions to overcome the bottleneck. This article describes the knowledge acquisition framework and how IC accelerometers can measure human motions. Empirical results that prove its effectiveness are also shown.