Polystar: Where Human Innovation Meets AI Execution
Discover why AI enhances, not replaces, developers, boosting productivity while relying on human creativity, direction, and technical expertise.
Discover the benefits of digital twins in telecom and how operators can use DataOps and service assurance data to drive performance and smarter operations.
Listen to this article
Digital twins have the potential to provide unprecedented opportunities to deliver real-time, operational insights that can drive performance improvement and support enhanced autonomous operations. But building a comprehensive digital twin is difficult. Can we secure benefits — today — by leveraging enhanced service assurance data, enabled by DataOps processing?
Digital Twins - Driving Insights and Operational Excellence
Digital twins are attracting considerable attention from a range of different industries. The basic concept is simple: they provide a complete, virtual replica of operational assets, which allows asset owners to experiment, model impact of changes, explore optimization.
For example, The Economist has noted how Digital Twins are helping diverse sectors such as manufacturing, aerospace and Formula 1, to name a few examples. Digital Twins, the author writes:
Will redefine what it means to run a company. Instead of coordinating disparate islands of automation, as is the case today, bosses will manage a constantly churning ‘flywheel’ fueled by data.
The Digital Twin Consortium (DTC) echoes this view, stating that:
By using a virtual model of an entire area or process, organizations can visualize and test out different initiatives, making data-driven decisions based on billions of network performance data points. These initiatives can then be evaluated through more precise enterprise-level analytics and location intelligence, to help identify optimal implementation scenarios.
How Can Digital Twins Impact Telecoms Networks and Operators?
That sounds very promising – but it could be applied to any industry or sector. So, what could this mean for telecoms networks and operators? Well, one of the roles of the DTC is to explore uses cases across sectors, essentially building templates and reference models, as well as best-practice guidelines.
You don’t have to look far to see how Digital Twins could benefit every kind of operator. For example, it notes that:
“Digital twins can simulate the propagation of radio waves in various environments and identify the optimal placement of antennas and repeaters for maximum coverage and signal strength. A digital twin of a satellite communications system or cellular tower can monitor its performance in real time and identify potential issues or faults before they become critical.”
Essentially, they not only provide an accurate representation of the physical assets deployed, but they also provide a playground for simulation. As such, it’s no surprise that telecoms operators understand the potential that they can offer.
Building a Digital Twin is, however, challenging. Why?
Digital Twins Will Require Data, Processing Capacity and Power
The answer lies in the need for data. If a model — or twin — is to be accurate, it needs sufficient data to enable meaningful analysis and for scenarios to be produced. Gathering all of the requisite data is clearly a significant challenge. Gaps would simply reduce the accuracy of the model and deliver only impartial results that would not be based on appropriate confidence levels.
As Cap Gemini has noted “the volume of data, the dynamic nature of networks, and the need for scalability pose considerable challenges in building network digital twins. They are dauntingly complex to create and maintain”.
If we consider that a live network includes multiple logical and physical domains, as well as different generations of technology, all running in parallel, as well as a growing range of services, we can see that building a twin of such a complex environment — even with all of the necessary data — would require significant investment in processing power and capacity, as well as sufficient manpower to maintain the model.
But comprehensive is one thing. We don’t necessarily need to model everything in order to deliver results. It may be sufficient to consider specific network challenges and to focus, first, on building models with the information we already have and exploring those in isolation.
Can We Make Use of What We Have to Begin Our Journey?
Sure, we may not have the 100% model we seek, but we may have sufficient to provide insights that drive benefits. So, if our ultimate goal is to implement and maintain a complete digital twin of our networks, is there a way to begin this journey today? Can we take incremental steps towards achieving this goal that can deliver benefits?
The answer is yes. That’s because operators are already deploying new data fabrics to ingest and manage data from a wider range of sources than ever before. With DataOps processing, disparate sources of data can be brought into a common foundation that supports a range of tasks and activities.
One key driver behind the adoption of DataOps techniques has been the need to enhance service assurance and to support a new range of complex services, particularly those enabled by 5G SA. To deliver on SLAs, meet customer experience expectations and to support dynamic, automated operations, we know that data is key.
As it happens, this same data can also be used to support early exploration of the live network through a nascent digital twin. We may not be able to explore everything, but we can explore some of the things of interest.
So, what we need to do is to consider what matters most to operations. Is it change management in the RAN? Is it services consumed by the highest value customer accounts? SLA performance for an MVNO partner?
The subjects of interest will vary from operator to operator but selecting those use cases that have the highest value (or potential impact) can be a way to target early-stage adoption of digital twins. You can isolate the necessary data and model just the
The expanded view of service assurance that is now available from Polystar’s solutions, combined with innovations in DataOps processing, mean that the data necessary for such explorations is now available.
By Bringing Together:
-
Network metrics – such as real-time control and user plane data from passive probes, vTAPs, as well as streaming data feeds from network nodes and functional entities
-
Performance management data – heterogeneous sources from different vendors and technologies
-
Telemetry data via file integration or SNMP
And adding other sources, like customer information from the CRM, the data held in network inventories, network speed tests, configuration data, and so on, we can isolate the information relevant to the domain or service of interest and build emulations that model what happens when variables change.
The Foundation Is Being Built – Prepare for Lift-Off
We have the foundation of a future, comprehensive digital twin but one that can be leveraged today to investigate specific challenges or scenarios. As we build on these foundations, we can begin to explore more scenarios and run the kinds of real-time simulations that will be the end goal of digital twin investments.
And this incremental approach is now gaining traction. For example, the Economist has also noted how leading manufacturers of jet engines are modelling the performance of their products – they are not looking at the whole airplane, but at an isolated (albeit crucial) component of the whole and they are not seeking to do so while in flight, which is a much more complex activity for which the industry is not yet ready.
This is where investments in advanced, AI-driven service assurance and analytics solutions that support operational excellence and new levels of automation can be leveraged.
In future articles, we’ll look at specific use cases and show how Polystar’s customers can take advantage of DataOps processing to begin the journey towards network and service digital twins – securing results today and preparing for a comprehensive approach in the future.
