Neural Group Implements Crowd Data Solutions to Combat Overtourism in Arashiyama

Neural Group Tackles Overtourism in Arashiyama

In a notable initiative, Neural Group, based in Chiyoda, Tokyo, has embarked on a project to collect crowd data to mitigate the effects of overtourism in Kyoto's picturesque Arashiyama region. This effort is part of a commissioned project by the city of Kyoto, aimed at addressing the surging visitor numbers, particularly during the autumn foliage season, when the area becomes extremely crowded.

In late November, coinciding with the peak tourist influx, the company installed AI cameras at the Togetsukyo bridge crossing. Traditionally, foot traffic was estimated using manual counters; however, the introduction of AI cameras allows for high-frequency data gathering, significantly enhancing the precision of visitor analysis. Considering the chaotic flow of people during peak times, Neural Group developed a new crowd detection technology utilizing head counting alongside the conventional cross-sectional traffic volume previously employed by older systems.

This advanced method has allowed them to gather more realistic data about visitor patterns. The insights gained from this project will be integral for discussions among stakeholders and in creating effective security and management plans in collaboration with local authorities.

Neural Group will continue to accumulate crowd data during key seasons such as spring outings and autumn leaf-viewing periods, further contributing to local tourism management strategies.

The Digital Flow by Neural Group

Expanding Implementation Achievements

Neural Group's Digital Flow service utilizes edge AI technology, ensuring high operational stability, accuracy, and robust information security. This service has successfully penetrated various sectors across Japan, including retail establishments, public facilities, and office buildings.

A Stable Operation with High Security

Neural Group specializes in developing unique technology that integrates advanced deep learning algorithms into compact edge devices. This method contrasts sharply with traditional approaches that rely on sending large amounts of footage to cloud servers, resulting in a more stable operation while maintaining high-security standards, protecting user privacy, and minimizing capital expenditures.

Easy-to-Use Interface

The web interface for administrators is designed to simplify data observation and accumulation. It allows diverse users to access real-time information through a straightforward interface, aiding daily operational needs effectively.