AI has pushed digital twins into the spotlight, they are the buzzword of the moment, and as with every new narrative in telecom, legacy systems are quick to rebrand themselves to stay relevant.
In telco we have seen this before. When Customer Experience Management (CEM) was the hot trend, every Network Performance Management tool claimed to be a CEM solution, even though they were still network-centric at heart.
When Cloud Native became fashionable, vendors relabelled old monolithic apps as “cloud native” just because they could run on VMware or in a container. I won't mention how many ‘Big Data’ applications ran on relational databases with some dashboards bolted on!
Now, the same cycle repeats. Legacy inventories give the appearance of digital twins, but look closer and you will see that it is just a thin layer, laid over old foundations.
Static Memories
One of the primary foundations of a traditional inventory is its data model. The data model is static, rigid and fundamentally costly to change.
In fact, it reminds me of one of those baby photo albums. Most new parents are likely to have had one of these albums, they had pre-printed slots for pictures or milestones; first word, first birthday, name of delivery doctor. In other words, a rigid schema, just like an inventory database, it tells you in advance what is supposed to exist.
However, reality does not always fit the template. If your child’s milestones did not line up, you would leave a gap or write something completely different under your own heading. The book looked neat on the outside, but beneath it was inflexible, locked to someone else’s assumptions.
Your data will always be unique, it may resemble a preconceived assumption, but it is never the same, hence why rigid templates inevitably force a workaround.
A Layer of Labels
Parents often tried to make baby albums more personal or tell a more correct and connected story by adding their own captions, scribbling notes or sticking photos that did not quite fit the template. It gave the appearance of something richer and more meaningful to them, ‘their story’, but the structure beneath was unchanged.
Legacy inventories have followed the same path. They add a layer of joined data, often through a labelled property graph, which makes them look more connected, more intelligent. In reality, it is an annotation on top of a rigid foundation.
A layer of labels can make navigation easier. It can even make the records look more modern, but it does not change the underlying truth. An inventory is a store of information which is difficult to accurately populate and even harder to keep up to date. Copying those imperfect records and storing them in a graph layer, is like taking a photo of a photo, each time you do it the quality gets worse and that ‘single version of truth’ drifts further away.
A semantic digital twin, by contrast, is built from the ground up as a connected model. There is no static assumption about what the data must look like. The inherent flexibility expects, and adapts to, the unknown specifics of your data. Relationships are defined by rules that dictate how the data links together, not by a predefined static schema. There is no predefined slot anticipating ‘First kiss’ or other spurious placeholders waiting to be filled. Instead, the model evolves with the data itself, storing only what is meaningful and relevant.
Your story may include photos from before birth, or snapshots at 3 months and 6 months. There does not need to be a predefined page or slot for that, it should be dynamically added with the rules creating the right connections to the rest of the story.
The Family Archive
Some of the most important photos do not live in your phone or the cloud. They sit in old albums or boxes in the attic. Faded baby snaps, grandparents wedding portraits, school photos taken decades ago. The quality may be poor but the moments are irreplaceable, they are part of the family story that exists nowhere else.
Inventories serve a similar role. They often hold legacy data that the live network cannot provide. site details, landlord information, patch panel configurations and splitter setups. It is important information to build a full end-to-end picture, it is data that does not live anywhere else. At times the data may be incomplete, inconsistent, and hard to maintain, but it is still essential. Data integrity and quality issues may persist, like the blurred focus of old photos, but you do not throw them away. They still serve a purpose, you still need that, often critical data.
What does not make sense, is trying to repurpose this store into something it was never designed to be. Moving towards a purpose-built digital twin, which will not dismiss this archive, but curate it, federate it with live data and most importantly put it into context so the story is whole, not partial.
The Digital Storybook
If the family archive photo album represents the past, the digital storybook represents the future. It is not the static album with fixed slots or fading prints, it’s a living evolving story.
New photos appear automatically, woven into the narrative as they happen. Faces are recognised, places identified, and events are grouped without you needing to curate them. Old scanned photos can sit alongside crisp new ones, but now they are connected, searchable and enriched with context. Patterns emerge, holidays in August, birthdays at the same restaurant, you and your best friend appearing together.
This is the essence of a true semantic digital twin. It is not bound by rigid templates or bolted-on labels. It is continuously updated, discovering connections, and reasoning about context and impact. It blends the legacy archive with the live feed to tell a story that is not only complete today, but continues to adapt as the future unfolds.
Building a Picture of Autonomy
Legacy inventories do not need to rebrand themselves as digital twins to remain relevant. They are relevant. They are a key function in the lifecycle of a network. They hold details of information that should not be lost, in fact the quality of that data should be a priority within every operator. In the world of automation, the value of an accurate inventory is priceless, just like those treasured snaps in a family photo album.
However, traditional inventories are static by design, rigid in structure, hard to maintain and often misaligned with the living reality they aim to represent. Adding a layer of labels or copying records into a graph may make them look modern, but it is an illusion, a photo of a photo, that will never carry the richness or accuracy of a true digital twin.
A true digital twin is something different, it is dynamic, capable of recognising connections and reasoning about meaning. It integrates legacy data together with live feeds, creating a complete picture that evolves flexibly as the network evolves.
The telecom industry has seen this cycle before, with CEM, Cloud and big data, and now again with inventories dressed up as digital twins. The lesson is the same, rebranding a legacy system does not make it fit for the next era. To move forward toward real autonomy, we must invest in technologies that are designed from the ground up to connect your infrastructure, reason over it, and provide a natural bridge to LLMs and Gen AI so that Autonomy can move from static slides to a living reality in your network.
