Furthermore, the two companies agreed to integrate Mitsubishi Heavy Industries’ highly scalable edge data center equipment technology with AITRAS, making it possible to execute AI in real-world environments, and aiming for the social implementation of new AI infrastructure.

■Background of the collaboration
In recent years, efforts to utilize AI to advance the operation and maintenance of equipment, known as “AI transformation,” have been accelerating in various industries such as energy and industrial machinery. In order to effectively utilize AI in such real-world environments, there is a growing demand for data centers that can be installed close to the operation site and provide an AI execution environment that combines stable communications and data security.

■Outline of demonstration experiment

To address these practical issues, as the first step in this collaboration, we will build an AI-RAN environment that utilizes “AITRAS” within the on-premises “DIAVAULT” solution. In this environment, we will introduce the AI ​​application developed by Mitsubishi Heavy Industries, “Defect Detection and Repair AI Application for Mitsubishi Heavy Industries Products,” and conduct demonstration experiments.

In the demonstration, image and video data sent from field workers’ smart devices will be analyzed in real time within a closed network environment, identifying faults and suggesting appropriate repair methods.

■Value gained from this demonstration
Through this demonstration, we will verify the following three points.

1. Real-time analysis using edge AI processing
   Since AI processing is completed within a closed network without going through the cloud, delays in data transmission and analysis can be significantly reduced under a stable communication infrastructure.
2. Enhanced security and closed network environment
   A secure AI inference environment has been established that prevents sensitive product data from leaking outside the company.
3. Efficiency and standardization of repair work
   AI-based fault identification reduces the burden on operations that previously relied on the expertise of skilled technicians.

■Future outlook
Based on this collaboration agreement and demonstration experiment, both companies will leverage their respective strengths to develop a scalable and reliable edge AI operations platform that continuously adapts to advances in AI. We will also promote initiatives to accelerate the “AI transformation” of customer sites by creating innovative solutions.

• 1A technology concept that realizes RAN (Radio Access Network) control functions and AI server functions on the same hardware platform. “AITRAS” is an integrated AI-RAN product that provides carrier-grade, high-capacity, high-performance, and high-quality communication networks, and also enables the provision of AI applications.
• 2Applications that perform data processing and AI inference where the data is created, rather than in the cloud.
• 3Small data centers located near users and devices.
• 4A system that allows servers to be located at locations such as companies, factories, and research institutes, rather than in the cloud, and allows data to be directly managed and processed.

• SoftBank, the SoftBank name and logo are registered trademarks or trademarks of SoftBank Group Corp. in Japan and other countries.
• Other company names, product names, and service names mentioned in this press release are registered trademarks or trademarks of each company.