1. background

Although the number of industrial accidents has decreased in recent years, and the number of fatal and injury accidents involving forklifts has also decreased, the number of fatal accidents in 2020 increased by 155% compared to the previous year. Forklifts have the second highest accident rate after trucks and passenger cars, and the number of forklifts equipped with drive recorders is increasing. By analyzing video captured by drive recorders, it is possible to provide safe driving guidance to forklift drivers, but many companies, especially in the logistics industry where forklifts are heavily used, are facing serious labor shortages. However, due to reasons such as not being able to handle video analysis over a long period of time, it is not being fully utilized for safe driving guidance for drivers.
For companies facing such issues, the Company have traditionally provided forklift operation diagnostic services that involve human video analysis, but since video analysis takes a long time, it takes a lot of time to provide the output. It required a lot of effort and cost.
With this background in mind, Sompo Risk and Macnica are collaborating this time to utilize the AI library of AI Research & Innovation Hub (ARIH* 1), Macnica 's AI expert organization, to transform drive recorder footage into AI. analysis, and succeeded in detecting dangerous driving footage with a high accuracy of over 97%. In addition, we have realized digital analysis of Sompo Risk's existing consulting, providing output in a short time, and reducing costs.
By providing this forklift driving diagnostic service, we aim to improve the driving skills of forklift drivers and contribute to the early detection of dangerous driving caused by cargo accidents and work-related accidents at companies, and the prevention of accidents.

2. Overview of forklift operation diagnosis service

(1) Flow of forklift operation diagnosis

Sompo Risk rents forklift drive recorders (manufactured by Data Tech: maximum of 5 units) for one week, analyzes the acquired video with AI, and extracts dangerous driving data such as sudden starts, sudden accelerations, and sharp turns. We summarize the diagnostic results in a report and provide it.

3. Cost/Period

① Expenses
・From 375,000 yen (excluding tax) per unit
* 300,000 yen (excluding tax) per unit from the second unit onwards
* 300,000 yen (excluding tax) per unit for vehicles with a Datatec drive recorder installed

② Period
1 week

4. Order target

We aim to receive orders for 50 units a year worth 18.75 million yen.

5. the next deployment

Sompo Risk will contribute to the prevention of forklift accidents through further development of AI analysis methods.

About Sompo Risk Management

Sompo Risk Management Inc. is a group company of Sompo Holdings, with Sompo Japan Nipponkoa Insurance Inc. at its core. We are developing "Risk Management Business" and "Cyber Security Business", and provide solution services such as enterprise risk management (ERM), business continuity (BCM/BCP), and cyber attack countermeasures.

About Macnica

Since its establishment in 1972, Macnica has provided cutting-edge semiconductors, electronic devices, networks, and cybersecurity products with added technological value. Based on our traditional strengths of global cutting-edge technology sourcing and technology planning capabilities, we are developing new businesses in fields such as AI/IoT, autonomous driving, and robots. With the slogan "Co. Tomorrowing," we will connect cutting-edge technology and Macnica 's intelligence to provide unique services and solutions, creating social value and contributing to the development of future society. the Company company is headquartered in Yokohama and operates global business with 85 locations in 23 countries around the world. Please see the website (https://www.macnica.co.jp) for details.

*1) About ARIH
We are Macnica 's AI expert organization whose mission is to connect cutting-edge technology and intelligence to create a happy future society for people around the world.
Click here for details: https://www.macnica.ai/research.html