Oleg Grinkevich
Oleg Grinkevich

Since its invention in the late 19th century, diagnosing what is wrong with a car has required a physical examination by a human being. This has long been an immovable fact of the automotive industry, particularly in the aftermarket parts sector. It underpins countless car businesses around the world. However, the emergence of AI technology could be about to change all that with the advent of 'digital twinning'.

The debate about AI's role in the automotive industry has been running for several years. There is little doubt that it will trigger major changes in how cars are designed, built, operated and maintained. Its impact here may prove outsized compared to many other industries. Twinning is one of the most fascinating aspects of this technological shift, opening up possibilities that were previously unthinkable.

In simple terms, twinning involves creating a detailed, data-driven digital model of a vehicle, built from real-time sensor inputs and performance data. This enables predictive maintenance rather than reactive repairs. The benefits are immediate and clear: less downtime due to earlier detection of issues; lower costs from fixing problems before they escalate; and potentially fewer emissions thanks to better overall performance. There appears to be little downside. The investment case reflects this: the global digital twin market in the automotive sector alone was valued at over $2 billion in 2024 and is expected to grow at a compound annual rate exceeding 30% by the end of the decade.

AI twinning also has the potential to improve vehicle safety. Being able to foresee problems before they occur carries obvious safety benefits. It is easy to forget — given dramatic improvements in safety standards — that deteriorating car parts still contribute to traffic accidents. Studies estimate that vehicle defects play a role in between 2% and 12% of all crashes, with brake failures and tyre issues among the most common causes. Human behaviour remains by far the dominant factor, but defective vehicle parts represent an area where technology could deliver a meaningful advantage.

Increasingly, cars resemble laptops and smartphones — not just because they are morphing into something you plug in, charge and go. Modern vehicles now generate, transmit and receive vast quantities of data in real time. A connected vehicle can produce roughly 25 gigabytes of data per hour, from engine diagnostics and tyre pressure to GPS positioning and driver behaviour. The rise of electric vehicles (EVs) has accelerated this trend further, given their reliance on battery management systems, software updates and continuous performance monitoring. This data ecosystem is what enables the creation of accurate AI twins.

Tesla already applies digital twin principles to battery management, using fleet-wide telemetry and remote diagnostics to monitor battery health and optimise performance over the air. Meanwhile, Siemens provides digital twin platforms widely used across the automotive sector, particularly in vehicle design, manufacturing and predictive maintenance systems. BMW has taken the concept further still: its iFACTORY initiative involves building entire production facilities as digital twins in a virtual space before a single brick is laid, reportedly projecting 30% savings in planning costs. It is increasingly difficult to imagine major car producers — and those involved in vehicle production or maintenance — not adopting such approaches in the near future.

Paradoxically, this approach could reduce new car sales in the long run. In Europe, the average age of passenger cars on the road has risen significantly in recent years, reaching around 12 years according to industry data — up from roughly 10.6 years in 2016. Tough economic conditions have played a role, but improving build quality and reliability have also contributed. People are simply keeping their cars on the road for longer.

It is easy to see how AI-assisted vehicle maintenance could reinforce this trend. A better-maintained, smoother-running car is less likely to need replacing. This is not just a question of how long a vehicle remains roadworthy; it is about how long drivers remain satisfied with its ride quality and performance. Put simply, AI twinning could enable cars to perform closer to their 'fresh off the lot' standard for longer, reducing the incentive to upgrade.

The ripple effects could be significant. Not only might new car purchases decline, but the aftermarket — an industry worth over €300 billion across Europe alone — will need to keep pace with technological change. Digital twins present a genuine opportunity here: by predicting when specific parts will fail, distributors and suppliers can optimise inventory, reduce waste and ensure the right component is available at the right time. Aftermarket distributors such as Armtek Group sit squarely in this space. Bluntly, if businesses are not already adapting to this industry-changing technology, they risk being left behind by those that are.