While, as detailed in the previous post in this series, private 5G will be an ideal building block for Industry 4.0, replicating what has been deployed by telcos is not the right direction for the enterprise. Private 5G networks should be optimized for enterprise deployments, must interface seamlessly with IIoT end devices, the edge cloud and the public cloud, as shown in Figure 1, and are best delivered as a service to minimize startup complexity and costs.
Most importantly, enterprise 5G should empower the creation of Industry 4.0 applications. Requirements for successfully doing so include use of cloud native technologies, the ability to integrate with the enterprise’s preferred hyperscaler and provision of edge cloud services.
Cloud native technologies empower organizations to build and run scalable applications in public, private, and hybrid clouds. Cloud native apps are designed and built to exploit the scale, elasticity, resiliency, and flexibility the cloud provides. The following features best illustrate this approach:
A cloud native approach offers a number of benefits when implementing Industry 4.0:
Parts of Industry 4.0 systems are inherently suited to run in a cloud. Taking machine learning (ML) as an example, storage and analysis of the massive amount of data used in ML training is a natural fit for implementation in an off-premise, hyperscaler cloud, while inference may need to run in the edge cloud close to the source of data in order to make more rapid and intelligent decisions.
Enterprise private 5G networks must support the interfaces required and integrate seamlessly with the major hyperscalers, ensuring that each enterprise can use their preferred hyperscaler. Empowering enterprises to pick and choose between hyperscalers enables enterprises to select the best cloud services for their business or use case.
Edge cloud services are an essential part of an Industry 4.0 system. An edge cloud processes data locally, with only select data backhauled to the public cloud. the enterprise can choose which data is stored locally in the edge cloud and which is sent to the public cloud. To continue the example above, while ML data storage and analysis is best operated in an off-premises cloud, applying ML models to react to new inputs should be done in the edge cloud, to minimize delay in the closed loop control system.
Processing data and storing data locally using an edge cloud provides the following benefits:
The edge cloud can be purchased as a service from a hyperscaler, can be deployed using a commercial offering such as VMware, can be implemented as a Kubernetes cluster running on bare metal using open source software, or can be purchased as a service from a private 5G service provider, a system integrator or a solution provider.
Private 5G will be an ideal building block for Industry 4.0 but duplicating standard telco deployments of 5G may not meet the needs of enterprises. Enterprise 5G must enable creation and deployment of Industry 4.0 applications. Requirements for successfully doing so include use of cloud native technologies including containers, microservices, service meshes, immutable infrastructure and declarative APIs, providing interfaces to integrate with the major hyperscalers so that enterprises can use their hyperscaler of choice, and provision of edge cloud services.
This post, the third in a series, is an expanded excerpt from the Ananki white paper “Enabling Digital Transformation and Industry 4.0 with Private 5G”. Register to view the complete white paper.
“What are Containers?” TechTarget
“Microservices” Martin Fowler
“Should I Provide a Declarative API?” meshcloud meshBlog