Automobile wheels are secured to a hub, a component that connects the rotating axle to the wheel, by multiple bolts and nuts. The nuts must be tightened to the specified torque using dedicated tools like torque wrenches. If the nuts are not properly tightened, they can easily come loose on roads, which presents the risk of major accidents. Particularly, if it happens to large vehicles like trucks and buses, given their large and heavy wheels, it can cause severe injury and damage to property. Even nuts properly tightened can gradually become loose due to vibra-tions generated by the vehicle. This calls for regular inspections to assess the status of the nuts not just after wheel mounting, but before using the vehicle.Shinichi Masunari of the AI R&D Group at the R&D Center explains the issue of wheel separation accidents involving large vehicles.“The number of accidents involving wheel separation from large vehicles has increased steadily over the past few years. Currently, the Ministry of Land, Infrastructure, Transport and Tourism, the Japan Automobile Manufacturers Association, and the Japan Trucking Association are working to promote awareness campaigns calling for thorough inspections. But how are these inspections actually done? Almost all rely on human perception. A mechanic will tap on a nut securing the wheel with a hammer, then listen to the sound and feel the vibration produced by the tap to determine whether a nut is loose or properly tightened. While anyone can recognize a nut coming loose and about to fall off, it can be hard to make a conclusive judgment if the tightening torque is just a little bit off. This made me think that we might be able to provide some support for these inspections. Seeing news of wheel separation accidents and hearing that these incidents continue to grow more frequent, I felt this was a pressing social issue that could happen to anyone anywhere. To solve this problem, I examined the feasibility of shifting the reliance on human senses to AI. Now we’re in the process of developing an AI-powered system to help with this pass/fail judgment.”The system under development consists of an impulse ham-mer equipped with a vibration sensor and an AI program that analyzes the vibrations.“The impulse hammer has a built-in force sensor. When a nut is struck with the hammer, the vibration transmitted to the hammer is converted by the sensor into electrical signals, which are sent to a computer on which the AI judgment program runs. The vibration data is analyzed and input into an AI model that determines whether the degree to which the nut is tightened is normal or abnormal.”Interview and article by Yoshinari Baba / Photo by Yuki AkabaIN THE BACKYARDOn the Front Lines of Technology DevelopmentGrowing numbers of accidents in recent years have been cases in which tires have suddenly come loose from large vehicles due to loose wheel nuts.Rion is developing an AI-enhanced system that assists pass/fail judgments of wheel nut fastening status by integrating artificial intelligence with our vibration measurement technology.We interviewed the engineer leading this development effort about the current status and issues they’ve encountered and his motivation for contributing to society through AI technologies.Accidents Involving Truck Wheel Separation: A Growing Social ProblemBehind the Scenes in Developing an AI-Powered 14Combining AI and Force Sensors to Prevent Wheel Detachment Accidents
元のページ ../index.html#16