An interview article about the Akihabara AI Security Project has been published.
- Apr 28
- 2 min read
An interview article about the AI security project we conducted in Akihabara—which was selected for the Tokyo Metropolitan Government’s “King Salmon Project”—has been published in “TOMORUBA.”
Article published on TOMORUBA：
In this project, we applied our crime and accident prediction AI, “CRIME NABI,” to the bustling environment in front of JR Akihabara Station to evaluate, based on data, the risk of accidents caused by pedestrian congestion and the effectiveness of security measures. By combining the analysis of pedestrian and vehicle data from AI cameras with the optimization of security personnel deployment using CRIME NABI, we confirmed that the system could reduce the mixing of pedestrians and vehicles by up to 37% while maintaining the same number of security personnel.
The article describes an initiative in which we utilized our crime and accident prediction AI, “CRIME NABI,” to address the risks associated with the mix of pedestrians and vehicles in front of Akihabara Station and Radio Kaikan. The initiative involved analyzing pedestrian and vehicle traffic using AI cameras, optimizing the deployment of security guards, and quantitatively evaluating the effectiveness of their interventions.
In this demonstration, we analyzed pedestrian traffic involving approximately 1.47 million people using nine AI cameras and implemented security measures based on a security deployment plan formulated by the AI. Verification using CausalImpact revealed that, compared to random security deployments, this approach reduced the number of pedestrians walking in the roadway by up to approximately 37%, confirming that AI-driven optimization of security deployments may be effective in separating pedestrians and vehicles and reducing the risk of accidents.
Moving forward, we will continue to collaborate with local governments, police departments, and private-sector companies to expand the scope of “CRIME NABI” beyond crime prevention to include safety measures for crowded areas, tourist attractions, event venues, and downtown districts. We are committed to advancing the social implementation of an AI-powered security model that simultaneously enhances safety and reduces security costs.
Related Information
Click here for news regarding project selections and details: https://www.singular-perturbations.com/post/_2628
Click here for the press release regarding the project selection and details: https://prtimes.jp/main/html/rd/p/000000048.000080699.html
