Work Pattern Analysis with and without Site-Specific Information in a Manufacturing Line

Takeshi Kurata, Rei Watanabe, Satoki Ogiso, Ikue Mori, Takahiro Miura, Karimu Kato, Yasunori Haga, Shintaro Hatakeyama, Atsushi Kimura, Katsuko T. Nakahira

Abstract: In recent years, there has been an increasing number of technological developments and practical applications for efficient and quantitative work analysis by measuring workers’ flow lines. However, there may be a delay in starting work analysis if it is started after preparing site-specific information, such as each worker's role and typical work areas. This paper reports on a case study of work analysis using geospatial intelligence techniques with and without such site-specific information. First, this paper introduces the work site targeted in this case study, the purpose of the analysis, and the data measured and collected at the work site. Next, it describes a series of methods such as indoor positioning, generation of work area transition model, clustering of work area transition instances, and exception extraction of the instances for the purpose of analyzing work patterns. A comparison of the clustering results with each worker's role and an analysis of non-routine work patterns are also reported.