A Japanese collaboration between Sumitomo Heavy Industries (SHI) and NEC has developed an AI system that automatically detects near-miss incidents using camera footage and sensor data from hydraulic excavators. This technology extracts, visualizes, and summarizes hazardous events from site-specific video and work logs, marking a first in automated safety reporting. SHI plans to integrate this system to enhance safety management and reduce accidents on construction sites.
What Happened
SHI and NEC created an AI system that analyzes video and sensor data from excavators to identify near-miss incidents. It automatically generates detailed reports, improving the speed and accuracy of hazard detection. The system is the first to combine time-dependent video footage with work logs for comprehensive safety analysis.
Why It Matters for the AECM Industry
This AI technology enables project managers and safety officers to proactively address safety risks by providing timely, data-driven insights. It reduces reliance on manual reporting, minimizing human error and improving on-site safety protocols. Contractors and engineers can leverage this system to prevent accidents, potentially lowering insurance costs and project delays.
What's Next
SHI Group will implement the AI system across its hydraulic excavator operations and plans to expand its use to other heavy machinery. Further development aims to integrate the technology with broader site management systems by the end of 2024.
Source: https://www.globalconstructionreview.com/japanese-team-uses-ai-to-spot-near-misses-on-site/. Read the original story →
