Phoxter to Exhibit at JIMTOF2026, Show All-Directional Inspection System With ‘AI Good-Product Learning’ That Requires No Defect Samples

Phoxter announced it will exhibit at JIMTOF2026 (Oct. 26-31), where it will showcase an all-directional AI visual inspection system integrated with a six-axis robot and featuring its new “AI Good-Product Learning” technology, which requires no defect samples. The exhibit will focus on a learning method based on good products and an inspection setup integrated with a robot. It will present visual inspection not as standalone image processing, but as a process coordinated with robot operation.

Phoxter’s announcement proposes reconfiguring inspection processes around learning from good products and integration with a six-axis robot, with implications for production-line setup and fixture design. If the company’s claims hold, the approach could make it easier to consider automating visual inspection in processes where defect samples are difficult to obtain. However, the company has not specified the technology’s range of applications or accuracy, so its practical effectiveness will need to be assessed through both the exhibit and evaluations of the actual system.

The new aspect of this announcement is its emphasis on training for visual inspection without defect samples, combined with integration with a six-axis robot. Rather than simply introducing image-inspection software, Phoxter is presenting data preparation and a robot process involving handling and pose control as an integrated system. This suggests that the focus is not only on inspection accuracy but also on reducing the burden of implementation on the factory floor. In a related article, Phoxter to Exhibit at International Logistics Exhibition 2026, we reported that the company planned to introduce its in-house AMR and low-code WCS. Its JIMTOF2026 exhibit can be seen as an expansion from logistics-related transport control into visual inspection on manufacturing floors. At the same time, both initiatives present systems that encompass operations, rather than hardware alone. AMRs, WCS and visual inspection are positioned not as standalone products, but as layers connecting processes on the factory floor. The industry-level implication is that competition in inspection equipment appears to be shifting beyond imaging and recognition algorithms to include the ease of preparing training data and implementation alongside robots. “No defect samples” could be significant for production ramp-ups where defective products are difficult to collect, but this announcement does not show how broad a range of products the technology can handle. Another point to assess is how far all-directional inspection with a six-axis robot can overcome constraints related to fixtures and workpiece shapes. The next questions to examine are the range of workpieces the system can handle, the amount of good-product data needed for training, inspection time and accuracy on the actual system. At the announcement stage, it remains unclear how far visual inspection automation can be made broadly applicable.

The primary source identifies the exhibit as an all-directional AI visual inspection system integrated with a six-axis robot. It also describes the learning method as “AI Good-Product Learning,” explaining that it requires no defect samples. The exhibition is JIMTOF2026, scheduled for Oct. 26-31. The published materials provide no further details on specifications, customers, unit numbers or terms of availability.

In Japan, automating visual inspection is directly tied to the manufacturing industry’s need to reduce labor, making the ease of deploying a combination of learning from good products and robot integration a key point of interest. A setup that uses a six-axis robot for all-directional inspection could be relevant to processes that involve significant burdens for fixtures or workpiece securing.