Our Japanese customer, MS&AD InterRisk Research Institute, just announced the start of an SHM-based traffic accident risk survey this October. The new service, named Collision Hazard AI Image Analysis, will focus on local governments and other organizations interested in deploying automated driving across Japan.

For the Japanese version, you can refer to the MS&AD website here. Several newspapers have published the news, including the industrial news daily Nikkan Kogyo Shimbun (translation below).

For Streetscope, this marks a significant milestone, amounting to the first scale deployment of our Streetscope Collision Hazard Measure (SHM) after a two-year collaboration with the team at MS&AD Interrisk. We look forward to carrying on helping our Japanese partners fulfil their vision for a safe, secure, and comfortable mobility society in order to reduce traffic accidents now and in the future.

_______________________________

Copyright: Nikkan Kogyo Shimbun

MS&AD InterRisk Research Institute to begin traffic accident risk survey next month, using AI to analyze dangerous spots

MS&AD InterRisk Research Institute (Chiyoda-ku, Tokyo; President: Masashi Ippongi) will begin a traffic accident risk survey in October to quantitatively show dangerous spots where “near-miss” accidents occur. The company will use image analysis technology for traffic conditions developed by an American company using artificial intelligence (AI) to visualize areas with a high risk of collision on maps. In order to maintain public transport in the region, there are expected to be more demonstration experiments for the social implementation of automated driving. The aim is to develop services using new technology and to win contracts for survey work from local governments and other organizations.

The new service is called “Collision Hazard AI Image Analysis”. It will utilize the video analysis technology of StreetScope (California, USA), which researches safe vehicle movement, to analyze images taken by drive recorders and images from fixed cameras installed at intersections and other locations. It will extract locations where near-misses are likely to occur based on factors such as vehicle speed, direction of travel and distance from other vehicles and indicate the risk of collision by a numerical value between 0 and 100, with a higher value indicating a higher risk of collision.

Until now, the primary way of identifying dangerous spots for traffic accidents has been through qualitative analysis based on people’s empirical rules and visual observation, so the Next Generation Mobility Group at MS&AD InterRisk Research Institute & Consulting, Inc. says that “the strength of this system is that it can show the results quantitatively.”

The company expects to use it for tasks such as surveying driving routes for use in automated driving verification experiments and evaluating driving routes in automated multi-storey car parks. The service fee varies depending on the length of the video to be analyzed, and for five hours, it starts at 3.3 million yen (including tax). The company aims to take on 30 contracts per year.

The Ministry of Land, Infrastructure, Transport and Tourism is working to expand the use of Level 4 fully automated driving under specific conditions in order to maintain local public transport and address the shortage of drivers. Public and private sector consortiums have been set up across the country to conduct demonstration tests of automated driving, and MS&AD InterRisk Research Institute & Consulting Inc. believes that the need for research work to assess the risk of traffic accidents will increase.

‍