It’s the final of a world sporting competition. A lot is at stake, including the need for highly reliable and scalable connectivity to provide the most enjoyable experience for attendees and viewers watching on televisions and other devices.
How do you ensure that you can deliver on that? Digital twins.
As a real-time virtual replica of a physical environment, digital twins can provide the ultimate testing ground to iron out connectivity issues long before any actual event gets underway.
Pre-event network planning and simulation, including crowd behavior, test connectivity scenarios, issue detection, and identifying ‘dead’ zones where no data can be up or downloaded, is possible by harnessing a digital twin that is connected to real data streams from sensors, networks, and devices.
With such utility, digital twins are emerging as a key decision support system to guide autonomous networks, reduce the risk of failures and improve outcomes in service operations. As the industry group, TM Forum, said recently: “Digital twin networks can help CSPs better understand the timeframes, processes and resources required for autonomous network implementation, risk of service interruption and reduce much of the uncertainty about the impact of AN use cases on the network.” (TM Forum, 2025)
Head of Digital Operations Portfolio Strategy, Cloud and Network Services at Nokia.
Introducing the service digital twin
Though network digital twins at the domain level have been around for a while, emerging standards are ushering in the exploration of service digital twins.
These are cross-domain and provide holistic models for understanding impacts on service performance, identifying potential issues due to changes in network behavior, and optimizing efficiency in operations. They offer a simulation environment for planning and operating networks while having a comprehensive view of network entities and their states, traffic, and interactions.
Digital twins also boost business agility and enhance customer engagement by enabling operators to be more agile in responding to market demands. They can also be used to optimize operational processes, thereby lowering costs and managing risk.
Increasing autonomy in operations – is there a catch?
Automation as a concept is not new in operations, but the evolution towards autonomous networks takes automation and AI to the next level.
AI in operations is about learning by leveraging data from different sources, analyzing what happened in the past to predict what might happen in the future, and using this in a preventative way to ensure that services never fail. While AI can predict the network behavior, it doesn’t predict the impact it will have on operations.
Another key element for autonomy is the use of GenAI, which brings in the capabilities to drive more system-led operations, like copilots and assistants, to make it easier to manage networks, auto-generating recommendations to aid decision-making and autonomously deriving resolution actions to fix issues.
While these enable higher levels of autonomy, some big questions still need to be answered around the control of autonomous network operations.
While humans are taken out of the loop to minimize errors, what is the risk of adding AI and GenAI into the loop? More reliance on intelligence coming from GenAI-powered agents carries the risk of potentially biased decisions, erroneous recommendations, or inconsistent outcomes.
Another big issue is AI maturity and how to ensure that the AI recommendations do not trigger adverse impacts on the network. With all these questions still open, there is a need to have a validation, or “watchdog” element, for autonomous networks.
Digital twins offer the solution
The simple answer lies in finding a mechanism to check the validity of recommendations, assessing the impact of actions on the network without directly impacting the network itself, and understanding the “what if” when services are created or fail, new actions are performed, or certain changes take place.
This is where digital twins come in by offering a simulation platform to accurately mimic the network, using intelligent models to assess the impact of network behavior, identify more efficient alternatives, and validate recommendations to achieve higher accuracy in network actions.
The digital twin can simulate various situations, derive the impact, and analyze what might happen to complement the capabilities brought by AI and GenAI. It can help to assess network resilience in different scenarios to identify capacity bottlenecks and manage dynamic service level agreement (SLA) definitions by considering the network behavior patterns to improve efficiency and optimize service delivery.
Most importantly, they offer a ‘production-network replica’ to validate insights, recommendations, and impacts on the network without having to use the network as the object of a trial-and-error experiment.
Real-world use cases of service digital twins
The service digital twin has several real-world applications, all of which contribute to achieving autonomous networks.
The first is service planning. A significant challenge in delivering cross-domain services is checking the availability of diverse resources across multiple domains and technologies. What makes it more complex is that failure in one area can lead to having to roll back the entire service operations, which can be costly, in terms of effort and financially.
Digital twins help predict service feasibility based on anticipated network behavior by leveraging AI-enabled prediction and service modeling capabilities.
Digital twins can also assist with service maintenance. By testing the twin with different data inputs and simulating varying network states and conditions, an environment identical to the real network is formed. This environment can be used to assess the impacts of network changes and maintenance actions on network services and even the end-user experience.
Finally, they can also assist with service operations.
As service digital twins model the service itself and hold a real-time replica of the service state and performance, they can be tested with various inputs to optimize service delivery and performance. This also means that AI-driven recommendations can be validated first on the digital twin before being applied to the network to check for adverse impacts. This is referred to as a “twin-first” approach.
Solving the autonomous networks puzzle
Digital twins are a key piece of the jigsaw puzzle of autonomous networks, which aims to drive network monetization, optimize customer experience, and enhance efficiency while controlling costs.
A “twin-first” approach allows operators to perform impact analysis and validate AI-driven actions before implementation, enhancing reliability and efficiency, and enabling zero-touch, zero-failure operations shaping the future of next-generation networks.
We’ve featured the best business smartphone.
This article was produced as part of TechRadarPro’s Expert Insights channel where we feature the best and brightest minds in the technology industry today. The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/news/submit-your-story-to-techradar-pro
https://cdn.mos.cms.futurecdn.net/PpMCMEBe3AE7VBSnGE6igF.jpg
Source link
