ABSTRACT
Nowadays, the number of tunnel accidents have been decreasing yeah by year and some advanced technologies such as AI and/or ICT devices are getting to be introduced into tunnel construction site. It might be true in near future that safety of tunnel worker is supported by such technology. In fact, we are now developing a “tunnel face monitoring supportive system” that helps crack identification on tunnel surface using a camera with AI. However, on the hand, safety at many tunnel worksites is still depending on workers attentiveness in the present circumstance. The work on the face of tunnels is carried out efficiently by a small number of sophisticated workers in Japan. In addition, it is no exaggeration to say that safety is also supported by their efforts. Behavior of workers at tunnel construction site are required multiple work at the same time, for example excavation of tunnel face, operate heavy machines, drill holes for blasting and monitor tunnel face for finding sediment collapse as soon as possible. Establishment of the effective “tunnel face monitoring supportive system” needs behavior analyses of tunnel workers and machines at tunnel face. Therefore, we analyzed and visualized their behavior using behavior-based safety procedure. As the results, we found that almost of all behaviors between workers and machines have not had time overlap. Based on these results, we are planning to continue developing the system that more effectively support the safety management of workers at the tunnel faces.
