China Mobile and ZTE Pilot Cloud SPN Computing-Network Convergence

press release: In September 2022, China Mobile and ZTE jointly completed the world’s first pilot project of Cloud SPN computing network convergence solution in Changzhou, Jiangsu. This innovative solution is mainly applicable to the scenario of upgrading traditional quality inspection to AI-enabled inspection in industrial production. This enables new algorithmic models of cloud edge collaboration and AI intelligent inspection to assist customers in automatic quality inspection of production line spray code. Add an AI board to his SPN product platform of existing commercial networks, load AI algorithm models to upgrade current product functions, and provide hyper-edge ubiquitous computing services.

Prior to this, some industrial parks had various “problems” in quality inspection. Lack of intelligent means, reliance on manual inspection, heavy consumption of human resources. Expensive dedicated devices introduced for automated inspection, many test sample variations, and inflexible customized algorithms for dedicated devices make it difficult to upgrade inspection systems. Poor accuracy of partially automated local inspection, real-time tuning of system parameters, and high service costs. The Cloud SPN Computing Network Convergence Solution by China Mobile and ZTE makes full use of the existing SPN capabilities to provide the park with local lightweight computing resources, combining edge computing, cloud computing and network functions. Integrate into edge cloud collaboration solutions. Adding an AI board to the SPN device and applying it to an industrial park to rapidly deploy local applications, the low latency of the transport network extends the public cloud to the edge to collect data in the park and provide real-time services locally. process. Using the results of our initial demand analysis and preliminary testing, we will validate how we can solve three key problems:

1. A low-latency metro transport network that supports fine-grained slicing private lines ensures secure data isolation for ToB application access to the cloud, extends the public cloud to the park, and moves public cloud intelligence to the park.
2. Production lines in the park are equipped with transport access devices that integrate machine vision chips and applications. Based on training samples, AI models can be quickly used for local real-time inspection, and user data can be stored and kept confidential at the factory.
3. A wide range of transport access devices can integrate computing power for ubiquitous deployment. The cloud edge collaboration architecture contributes to lightweight edge computing, and the intrinsic computing power of network devices greatly reduces the deployment and O&M costs of hyper edge computing services.

For example in a technical quality inspection scenario. A specific controller and industrial camera work with her PLC (Programmable Logic Controller) for original on-site quality inspection. There are some issues with this process. There is no service data association between quality inspection workstations on the production line, requiring a specific set of controllers and industrial cameras to be deployed on each workstation individually. Controllers become complicated by quality inspection processes or changes in standards, making them difficult to upgrade. The controller agent’s service response capacity is not keeping up with the requirements, and the service localization is insufficient. Cloud SPN Computing – After the pilot trial of the network convergence solution started, ZTE HEC (Hyper-Edge Computing) devices (SPN access device 6180H with integrated AI board) were deployed in the production line, and the quality of the spray code was confirmed by the factory. will be inspected. Create an eighth workstation to conduct joint debug testing and address customer requirements for spray code. This is the first time SPN devices provide local intelligent computing services to industry customers. Close cooperation will continue in the second phase. One of his HEC devices can gradually take over the inspection tasks of his seven other workstations on the production line. Benefits of the overall solution are:

1. Data between processes can be centrally analyzed and managed, and in the future new services such as prediction of abnormalities in all processes of production lines can be provided.
2. Even if processes or standards change on your workstation, the software and algorithm models can be trained to make new inspection procedures readily available, facilitating upgrades.
3. In addition to ZTE’s advantages in localization services, models can later be trained on mobile clouds through edge cloud coordination and inference can be done on local SPN devices. The model is delivered to his HEC device in the factory via SPN MAN channel and can be loaded and updated online.
4. Cloud SPN computing network convergence solution reduces costs than previous industrial park controller solutions.

This solution upgrades traditional quality inspection to AI-enabled precision inspection. Leveraging a wide range of SPN devices invested in China Mobile’s existing network to generate a new model of edge-cloud collaboration for innovative applications, verify the feasibility of hyper-edge ubiquitous computing applications, and It sets the foundation for people to boost. ToB service with reproducibility of scenarios.

Subscribe to our daily newsletter Back