It is necessary to increase the productivity in every shipyard. To visualize the actual work status in any industrial activity, the observation of the work being performed is essential. However, current work observation requires both time and labor and some cases, shipyards hesitate to implement work observation. The aim of this study is to develop a new work observation method by use of deep neural networks to reduce the disadvantages of current work observation approaches while improving the accuracy of work identification.

Keywords:
neural network, identification ability, identification probability, dnn work observation, machine learning, deep learning, welding work, air chipper, checking welding arc, shipyard
Subjects:
Information Management and Systems, Artificial intelligence, Neural networks
This content is only available via PDF.
2018. The Society of Naval Architects and Marine Engineers
You can access this article if you purchase or spend a download.