.png){kind=logo}
Newsletter N°86 - February 2026
📶 Smart Industry / Robotics & AI: NTT and Partners Validate Remote AI-Powered Factory Inspections Over IOWN APN
NTT COMWARE, NTT DATA Group Corporation, ITOCHU Techno-Solutions Corporation and Mitsubishi Chemical Group Corporation have jointly completed a verification experiment demonstrating remote-controlled robotic inspections using the IOWN All-Photonics Network (APN). Announced on January 20, 2025, the initiative aims to reduce the operational burden and safety risks associated with factory maintenance, particularly in large-scale industrial environments.
The experiment leveraged the IOWN APN, an ultra-high-capacity, low-latency optical infrastructure, to transmit real-time 4K 60fps video over a 120 km network between Odaiba and Gotanda. This enabled operators to remotely control inspection robots while simultaneously running AI-based video analysis to detect structural anomalies.
Two key inspection capabilities were validated:
-
Real-time crack detection on wall-embedded pipes using image recognition AI (Deeptector), with results visualized instantly within a digital twin environment powered by NTT COMWARE’s 4DVIZ technology.
-
Precise pipe vibration analysis, conducted by NTT DATA using a remotely operated Unitree Go2 robot. The system successfully extracted amplitude and frequency data from video streams, targeting practical industrial benchmarks such as 0.1 mm amplitude and 60 Hz vibration frequency.
Image of initiative's labor-saving approach for factory inspections
Importantly, the verification confirmed that latency remained within practical operational thresholds, with transmission delays proving less impactful than device-level processing delays. ITOCHU Techno-Solutions further demonstrated that using RDMA-compatible FA cameras reduced CPU load by approximately 6% compared to TCP transmission, highlighting potential efficiency gains in large-scale deployments.
From an industrial perspective, the project addresses structural challenges in manufacturing maintenance:
-
Time-consuming inspections across vast facilities
-
Safety risks during high-altitude or hazardous checks
-
Variability in human judgment when assessing deterioration
​
By combining remote robotics, high-fidelity optical networking, AI-driven analysis, and digital twin visualization, the consortium validated a pathway toward smart maintenance and predictive infrastructure management.
​
All four companies are members of the IOWN Global Forum and positioned this initiative as a reference implementation model for remote robotic inspection use cases. The next phase includes establishing a communication environment at a Mitsubishi Chemical manufacturing site to conduct further anomaly detection trials under real industrial conditions.
Longer term, the partners aim to expand toward multimodal AI analysis, simultaneously integrating video, sound, and other environmental data streams from multiple robots to achieve real-time, high-accuracy remote situational awareness.
The experiment reinforces IOWN’s positioning not only as a next-generation telecom infrastructure concept but as a foundational enabler for Industry 4.0 and smart factory transformation.
Image of factory inspection with smart robots