Experts believe that digital twins have the potential to revolutionize mobile networks, and will play a critical role in amping 6G’s performance
A digital twin is a virtual rendering of a physical or digital object, so accurate that it reflects the behaviors, peculiarities, and nuances of the object down to the last detail. If the real-world object has any grand features, it will have them; if it has defects, it will have them too.
“Digital twins aren’t just animations; they’re high-fidelity simulations,” explained Siemens’ CTO and CSO, Peter Koerte, in an interview with Wired. “Powered by AI, they don’t just simulate possibilities—they analyze countless scenarios, identifying the most likely outcomes and optimal solutions.”
At a high level, a digital twin is made up of five core components — a mathematical model or the virtual representation of the real-world object or system; data sources that it pulls data from; a feedback system that creates a bi-directional connection between the real object and the model enabling continuous tuning and pruning; an analytics engine — often powered by AI/ML — to make sense of the input data, simulate future scenarios, and run predictive analysis; and lastly, an UI to publish the output for teams to see.
There are broadly two kinds of digital twins that are in use today — one that replicates physical products or components, and is used in product development; second, that replicates systems and processes, and is used in behavior prediction and lifecycle monitoring. Both kinds rely on real data to accurately mimic and optimize the real version.
Uptake across industries
A 2023 market research found that approximately 75% businesses invest in digital twin technologies in some form or fashion. Deployment has risen across industries in the past couple years as research shows its potential in anticipating scenarios accurately and in a fraction of the time of regular prediction tools.
A a disruptive force powering Industry 4.0, the technology is transforming high-end industrial complexes, its use often spanning quality control in Smart Factories to drug manufacturing in pharmaceuticals. More recently, the technology has also found use in many groundbreaking initiatives for developing a model of the sun — a project that IBM and NASA co-developed, or creating a digital clone of the human brain — or the human heart, and so on.
Now one upcoming technology that is poised to revolutionize sectors from industrial automation to telemedicine, autonomous transportation to defence communication and AR/VR needs a boost of digital twin to deliver on its promises.
Digital twins for 6G
There is a growing consensus in the industry that digital twins can unlock new use cases for 6G networks. 6G represents a significant evolution in wireless connectivity. It has been described as the springboard for advanced technologies, like metaverse, VR/XR, autonomous vehicles, smart cities, and industrial IoT. Use cases for 6G networks range from schools and academic institutes to hospitals, tourism, enterprises and data centers. But, to take it to the next level where operators can deliver improved resiliency and always-on connectivity without breaking a sweat, they need digital twins.
You see, a network digital twin can mirror the complex 6G ecosystem right from its physical infrastructure layer where network devices live to network operations all the way up to the software plane. It can simulate the dynamic interactions and dependencies, the states and operating conditions — and fill gaps in data with synthetic data creating forward-looking demand and failure scenarios. And this is groundbreaking for network management.
What-if scenarios: One thing a digital twin is especially good at is generating hypothetical scenarios. Using a combination of historical and real-time data, operators can use it to generate various scenarios and measure key performance indicators to tally with predefined standards. This allows operators to not only take measurement of the network performance at present, but also explore a multitude of “what-if” scenarios that may happen. The best thing of all, the scenarios can be orchestrated and studied in a completely risk-free sandbox environment without ever touching the real production network. Additionally, it can also outline future performance trends that is helpful to know what to expect and avert crisis.
AI training and inference: 6G is often described as AI-native connectivity. Explaining what that is in practice is however a bit complex. Blue Planet’s VP, Kailem Anderson, shined light on it during an earlier interview with RCR Wireless News. He said,“[AI] will be built into the chipsets, the hardware protocols, the software stack and various abstraction layers so that the network is truly intelligent. What does that mean? It means 6G will truly embrace principles around self-healing, self-optimizing, self-organizing, so that the network operates in a truly declarative or intent-driven way.”
This is one of the reasons why digital twins are relevant in the context of 6G. AI is all about good data. The data output from a digital twin can be highly valuable for training AI/ML algorithms in 6G networks. The context-rich information can not only keep the models updated, but also enable accurate inferencing which is essential as mistakes can cost millions.
Power optimization: With each new generation of wireless standard, the energy cost climbs higher. Same goes for 6G. 6G’s extreme speed is accompanied by heavy energy demands — especially as AI is at the core of its vision. That increases its environmental cost while introducing complex grid management for many aging infrastructure.
Once again, digital twins can help optimize the power draw of 6G networks. By providing real-time insights on traffic patterns and energy usage during peak- and low-demand periods, and simulating future scenarios, it can enable smart power management. This can help operators predetermine demand variations and work out the most energy-efficient configurations to ultimately, reduce energy usage and carbon footprint. With network AI-powered digital twins, operators can also dynamically turn on and off network resources, based on the hour of the day and traffic load, saving energy when possible.
