Image recognition technology to visualize unsafe behavior
Helping to improve the on-site environment by monitoring workers’ safety
Stable operation of production sites is important for companies in the manufacturing industry, and occupational safety is one crucial factor.
However, manpower shortages have resulted in insufficient resources to share and manage know-how for safe work. A solution utilizing this technology visualizes the status of compliance with rules while monitoring workers in order to support DX and create a safe workplace.
We have developed a technology to determine actions that violate rules and cause unsafe behavior, by analyzing images captured by an IP network camera (our MOBOTIX). The skeleton detection technology of Konica Minolta’s FORXAI Imaging AI is used to detect a worker. It determines the person’s position and detects actions that could cause the worker to become trapped or caught, such as extending a hand near a machine or entering an off-limits area, as events. Alerts calling for action are issued in the worksite in real time, thereby helping ensure safety. A dashboard allows managers to easily check the images of events and the number of violations of rules. It presents data for studying measures.