Data-Driven Transformation in Telco Operators | White Paper | Polystar
The Telecom Industry is facing a deep transformation. Read about the CSP status, as well as the drivers for transformation. Download here
Blog
20 October 2025
Author
Automating Service Assurance is a critical step to drive operational performance for telecom operators – challenged by emerging 5G SA services.
Blog
20 October, 2025
Author
Predictive Algorithms Enabled by AI and ML
This approach ultimately depended on users taking actions – hunting down, for example, the root cause of dropped calls in a given cell site and then initiating the appropriate remedial actions. Of course, increased use of ML and AI accelerated this process, leading to the arrival of more predictive algorithms – so that issues could be identified before they caused any impact.
5G SA Brings New Demands for Service Assurance
All well and good – and excellent signs of the progress we have made. But, with 5G Standalone, that’s not going to be sufficient. If we have services that depend on the highest levels of performance, predictions will help, but post-facto remediation won’t cut it.
If we then consider the increased volume of data available for any given service, we can see that the task of finding needles that point to issues will only get harder. The haystack is growing. Rapidly – and so too is the range of services delivered.
Automation is Critical
These factors lead to the inescapable conclusion: automation is the key factor for delivering service assurance in the new era of 5G SA enabled communication. Of course, automation can already help.
With our dropped calls case, once we’ve identified where the problem is occurring, who’s affected and drilled down to the root cause, we have solutions in our toolbox that we can apply. Automation allows these to be attempted with the best match for the root cause identified.
Edge Processing is Just One Driver for Automated Operations and Service Assurance
Now let’s consider a 5G SA service that requires edge processing – for an automated industrial process. In this case, the UPF (User Plane Function) is distributed and positioned at the edge of the network, close to the source of activity, the factory in question. That’s because the manufacturing operations (in our hypothetical case) involve industrial robots and require a high degree of precision, with real-time feedback. The control is enabled wirelessly, which means that the applications that drive production depend on low latency performance.
They are also subject to strict safety requirements – so there’s no margin for error. That means that, if something goes wrong, then failsafe and protection mechanisms need to be applied. So, while the data that controls the onsite applications is transparent, the performance of the mobile network is integral to ensuring effective, safe and secure operations. These requirements change the game for service assurance.
Low Latency Brings New Benefits and Challenges
In practice, this means that the connectivity from the RAN to the gNB (gNodeB), and then to the UPF and 5G core needs to be assured to deliver the performance required – which, in turn, means that any errors that disrupt the time critical communication need to be identified (or predicted) before they impact the services delivered.
Given the millisecond performance required, it’s simply not tenable to assume that human oversight can ensure the remediation takes place before there is an impact – which, in a manufacturing context could have a significant cost in terms of lost production.
While such services can be delivered by private networks, dedicated to a specific purpose or campus location, they can also be enabled over public networks. In either case, performance is paramount – but in the public case, there will also be many other services delivered in parallel, adding further complexity to the task of delivering assurance.
Of course, not every service has such safety requirements – but many 5G use cases do share these demands. And, even if the impact is a minor disruption to a consumer service, with a multiplicity of such services running in parallel, there will be just too much for traditional operational approaches to cope.
