NTN, Yazaki Corporation Test Product Inspection System Combining Robotics and AI

NTN and Yazaki Corporation announced on November 5, 2026, that they would collaborate on product development using robots and AI. As the first project, they will test an inspection system that combines NTN’s robotics technology with AI image-processing technology developed by Yazaki Corporation to inspect product appearance and quality. Testing will take place in Susono, Shizuoka Prefecture, where Yazaki Corporation’s research and development site, Innovation Hub-REN (Ren), is located, using products that the company actually inspects.

According to The Nikkei, the companies will use products that Yazaki Corporation actually inspects to assess the system’s effectiveness and deployment potential. Moving beyond a demonstration to evaluate its application to real products will help determine whether it can be used on manufacturing floors.

The focus of the collaboration is not only on combining robotics and image processing, but also on testing the system’s effectiveness and deployment potential using products that are actually inspected. Demonstrating operation and determining whether the system can be applied to appearance and quality inspections of specific products are distinct steps. Whether the results lead to adoption in actual manufacturing processes will need to be assessed, including in terms of the products inspected and the evaluation methods. NTN demonstrates AI-based visual inspection described a visual inspection demonstration using a different combination of technologies. The current collaboration has been reported as testing Yazaki Corporation’s AI image processing with actual products in the same inspection field; it has not been confirmed that it uses the same system or technology as the earlier demonstration. Yazaki Corporation has also announced a development platform for manufacturing sites in Airion, Yazaki Corporation announce collaboration on physical AI development, but no connection or division of roles between that effort and the current system testing has been disclosed. The initiatives should therefore be considered separately, rather than treated as a single system, with their respective target processes and technology configurations kept distinct. From an industrial perspective, the test will assess whether a process combining machine movement and recognition can work, with a robot operating in relation to the inspection target and image processing judging the product’s condition. NTN’s robotics technology has been reported to feature smooth movement using the company’s bearings, and to be usable for greasing and cleaning as well as visual inspection. However, those tasks have not been identified as part of the current testing. Key issues include whether the companies will report inspection accuracy, processing time, the range of products covered and conditions for deployment at manufacturing sites alongside the evaluation results. The system is at the stage of testing its deployment potential; the effort does not confirm adoption in mass-production processes or the realization of autonomous control. See The Nikkei’s report for details.

This is an example of a Japanese company testing the potential application of product inspection that combines robotics and AI image processing at a research and development site. However, the results and adoption in mass-production processes have not yet been confirmed.