Towards an industrial internet with AI-enabled service assurance

The connectivity business is changing and service assurance needs to evolve with it if the vision of an industrial internet – cloud-based interconnectivity of IT systems across multiple cloud providers is to become a reality.

Rather than connectivity being delivered over single networks, in the future it will use many different physical network elements such as 5G, mobile private networks, edge, hybrid clouds, SDNs and VNFs. These networks and network elements will be exposed via network APIs into the services layer, allowing service providers to deliver differentiated experience to their enterprise customers. CSPs will have the possibility of becoming “outcome providers” in the industrial internet.

But delivering mission-critical enterprise-grade services is going to require new service assurance capabilities.

Systems integration specialist Cognizant has partnered with ServiceNow to develop new service assurance solutions. As part of the partnership the two firms have broken the service assurance value chain into four value chain elements or processes:

• Experience assurance. The customer wants a seamless experience across any app, any data in the cloud and wherever they are.
• The next layer down is orchestration assurance which involves managing the end-to-end delivery of the service.
• Service assurance which is specifically about managing the network
• Data assurance which gives an end-to-end view of the service that is being managed.

There are many dependencies for CSPs that wish to take such a holistic approach to service assurance. First, the CSP needs to have a system of record in place that is unique for customer service and network management. Data must be centralized and collaborations tools introduced to allow different teams to work collaboratively. A knowledge platform is needed to ensure that this collaboration is effective. By deploying a single pane of glass different teams will have full visibility into network and service performance.

AI will play an increasingly important role in fulfilling CSPs’ service assurance requirements. End-to end network and service automation will ensure that network and service issues are addressed in a time manner with no need for manual intervention. As CSP adoption of AI matures, the discussion will move onto whether the network itself can make operational decisions to precision new network capacity as and when there are surges in demand.

The vision of autonomous networks remains just that – a vision. In the short to medium term, and to prepare the foundations for network automation and a new approach to service assurance CSPs must address some of the more fundamental issues within the network and operations systems including inefficient processes, business and technology siloes and building new ecosystem orchestration capabilities that give them foundational capabilities of delivering the connectivity required to support an industrial Internet.