NTT docomo began a demonstration experiment on Friday, the 22nd, to detect bear sightings in real time by installing surveillance cameras linked to image recognition AI at base stations in Hokkaido.

A stable mobile network is established around the base stations, enabling real-time video transmission and rapid notification even in rural areas. Furthermore, utilizing existing equipment minimizes environmental impact and implementation costs, and the ability to monitor vast areas in the future is another advantage.

The system analyzes camera footage installed at the base stations using image recognition AI to detect bear sightings. To avoid missing urban bears that may appear day or night, the system aims for stable detection performance even under environmental fluctuations such as lighting conditions. By running the high-speed inference-capable image recognition AI on the low-latency docomo MEC, real-time detection is made possible.

This initiative utilizes docomo’s communication facilities and AI technology to balance the preservation of the natural environment with the safety of human society, aiming to contribute to the realization of “nature positivity.” It is also part of the efforts outlined in the “Mid-Term Roadmap for Biodiversity” formulated by the company.

※Translating Japanese articles into English with AI