Perception of safety vest and helmet of field workers in automotive sector with artificial neural network method within the scope of digital transformation

Keywords: Digital transformation, Work safety, Personel Safety Equipments, Artificial Neural networks, Yolov3 Algorithm The aim of this study is to minimize work area safety risks with artificial intelligence supported systems by inspecting work safety tool use such as personel safety equipments like hard hat and work safety vest in dynamical work enviroment at automotive industry. Use of personel safety equipments is fundemental for avoiding risks of losing labor and further more losing trained human resource wich is most important for automotive industry. Current studies shows that personel safety equipments are avoided because of personel reasons such as work comfort and lack of awareness. And it is found that inspecting/auditing is very important in order to promote use of equipments. With scope of this study, image data samples that are downloaded using Google batch Data image downloader and real online images that are taken with IP cameras are labeled and will be used with model using neural networks that are trained by Yolov3 neural network algorithms in order to inspect use of personel safety equipments. Scope of this this application is limited to inspecting hart hat and work safety vest use that can be installed at automotive industry production lines or observe worikg personel at field in order to feed back to safety managers. With the potential of this application it is possible to give ability to control the process by making automated line stop in case of determining personels that doesn't use hard hat and work safety equipments or determining enterence of this personel to line.