SoftBank Corp. and Mitsubishi Heavy Industries, Ltd. have launched a partnership in order to explore the new applications when it comes to edge AI in the edge data center domain through leveraging the AI-RAN product -AITRAS from SoftBank
Under the first initiative when it comes to this collaboration, both companies are going to conduct a field trial of edge AI applications through making use of AITRAS at Yokohama Hardtech Hub.
The field is going to take place in the Edge Data Center at MHI, which is installed on the YHH premises. Via this initiative, the companies are going to verify effectiveness in terms of high-speed AI inference delivered through secure as well as stable communications within an on-premises environment, which is isolated from external networks.
SoftBank, as well as MHI, looks forward to creating new value as well as use cases by way of AI-RAN while at the same time promoting the development, along with social execution, of the next-gen AI infrastructure.
It is well to be noted that in the recent years of AI transformation, the use of AI in order to advance operation as well as maintenance of facilities has been speeding up throughout many industries, such as energy as well as industrial machinery. AI utilization within these fields needs an environment that balances the stability of. communication in addition to data security. Specifically, there is a growing demand that is seen for systems that can go ahead and complete the AI inference at a local level at the customer sites or on-premises without depending upon the cloud.
In order to address these issues, SoftBank, along with MHI, is going to, as the first phase within their collaboration, build AI-RAN product -AITRAS from SoftBank within the edge data center of MHI, which will be located on the premises of YHH from MHI. Both the companies are going to conduct a field trial by operating the Defect Detection & Repair AI Application for Mitsubishi Heavy Industries Products by MHI within this environment.
In the field trial, the Defect Detection & Repair AI application for Mitsubishi Heavy Industries products is going to evaluate images pertaining to defective MHI products sent by the field workers through the smartphones in an instant. The trial is going to verify the ability of the application to identify the defective areas and also gauge the right repair methods.
Value when it comes to the field trial
Via this field trial, the following three values are going to get verified –
Real-time analysis via edge AI processing
Through completing the AI processing in an on-premises environment that’s isolated from the external networks that are independent of cloud connectivity, the system prominently decreases the transmission of data and analysis latency while at the same time making sure of communications that are stable.
Realization of high-level security along with closed-network operations
The environment helps to secure AI inference sans transmitting the confidential product data outside of the network.
Enhanced efficiency and standardization pertaining to repair operations
Through identifying the defective areas via AI, the system reduces the operational burdens that have in the past been dependent on the expertise of technicians who are skilled.
What lies ahead
Through leveraging respective strengths when it comes to both the companies’ technologies and expertise in AI-RAN and MEC – Multi-access Edge Computing of SoftBank and data center facilities as well as engineering technologies of MHI, the companies look forward to building a scalable as well as highly expandable edge AI platform that can continue to go ahead and adapt to the AI evolution. Moreover, both the companies are going to promote the initiatives in order to create novel solutions, which in a way speed up the AI transformation at the customer sites.