Ageing Infrastructure Inspection

Overview

• Japan is dealing with aging infrastructure and a shrinking workforce, necessitating an estimated 195 trillion yen (approximately 1 trillion GBP) for maintenance over 30 years, against a backdrop of local government financial challenges. The lab is collaborating with municipalities and focusing on disaster prevention and community safety, working with Mind Foundry to utilize AI in infrastructure maintenance. We are developing an AI-based tool for efficient infrastructure inspection, with ambitions for a nationwide rollout and broader applications in infrastructure upkeep and related risk management services.

​

Detail

Background​

• Many of Japan's social infrastructures, such as bridges and tunnels built during the high economic growth period, are nearing the end of their lifespan. The Ministry of Land, Infrastructure, Transport, and Tourism estimates that approximately 195 trillion yen will be required for maintenance and management over the next 30 years. However, many local governments face serious financial shortages for maintenance. Additionally, Japan is experiencing an aging population and a declining workforce.

• Aioi Nissay Dowa Insurance has formed comprehensive collaboration agreements with about 500 local governments across Japan to contribute to regional revitalization. Their efforts mainly focus on disaster prevention and mitigation, as well as ensuring safety and security in local communities and everyday life.

• Mind Foundry, a founding partner of AIOI R&D LAB-OXFORD, has experience in AI development for aging infrastructure in the UK and is focusing on this sector.
   

Hypothesis

• In many local governments, the use of AI for infrastructure maintenance and management is not yet widespread. By combining Mind Foundry's AI technology for aging infrastructure with Aioi Nissay Dowa's risk management capabilities and network of local governments, it becomes possible to maintain and manage infrastructure more efficiently and effectively with fewer personnel.
   

R&D

• After discussions with Susono City in Shizuoka Prefecture, a comprehensive partnership agreement site, we have begun the joint development of an AI-based bridge inspection support tool.

• This tool is being developed with the goal of reducing the workload of inspectors and standardizing the assessment of infrastructure health. It utilizes AI to automatically detect and quantify damage from photographs taken by inspectors and to categorize the health status of the infrastructure into four divisions as defined by the Ministry of Land, Infrastructure, Transport, and Tourism.
   

Next Step

• We plan to start implementing the beta version of the tool in Susono City from July 2024, with further refinements to be made within the year. Subsequently, in January 2025, we aim to roll it out to municipalities nationwide.

• The AI will be enhanced by utilizing data such as vehicle vibration data obtained from telematics data. We are also working on applying AI in the maintenance of infrastructures other than bridges.

• Additionally, we are considering the development of associated risk management services and insurance products.