AI in the repair shop: what the data actually shows

Search "AI car diagnostics" and the same number keeps coming up: an automated inspection system supposedly detects vehicle defects with 96% accuracy, against just 24% for a manual check. That's the figure published by UVeye, an Israeli company that sells these systems to dealerships and large fleets, including Amazon and General Motors. Before treating it as a benchmark for your own shop, it's worth checking where it comes from.

The most-cited number has no independent verification

The official press release announcing that Amazon uses UVeye's technology to inspect its vans, distributed via PRNewswire, doesn't contain the 96% and 24% figures at all: it talks about automating inspection at scale, without stating accuracy percentages. Those two numbers appear only on UVeye's own site and marketing materials, later repeated by trade outlets without independent checking. No academic study, certification body, or independent measurement confirms them: it's a company describing the performance of the product it sells, not a figure validated by a third party.

There's also a question of scale and context. The 96% figure refers to automated scanning lanes installed in hundreds of dealerships and logistics centers, processing millions of vehicles a month, mostly to catch defects in tires, undercarriages, and bodywork. It's an industrial setup built for large fleets, not a tool designed for the mechanical diagnostics of a small Italian repair shop. Even if the number holds up in its original context, applying it to a neighborhood workshop is a leap the available data doesn't support.

What the Italian data actually shows

A less flashy but verifiable number concerns hiring. In 2024, Italian businesses sought 39,000 car mechanics and repairers, and 52.9% of those openings required the ability to apply digital technologies. The figure comes from the Excelsior information system run by Unioncamere and Italy's Ministry of Labour, cited by Confartigianato Imprese in a report on the auto supply chain published on 5 September 2025. The same report found that 72.6% of the openings also required skills in low-environmental-impact products and technologies: a sector asking its workforce for new skills on two fronts at once, digital and environmental, more than the current conversation about AI in the repair shop suggests.

It's a figure about the people working in a shop before it's about the tools a shop buys: a business hiring or training staff today already finds, in the sector's own labor market, widespread demand for digital skills.

In Sicily the investment signal is positive, but it's regional

On the investment side, Confartigianato Sicily's MPI Observatory released a regional figure in late May 2026, at the AutoServiceTec trade fair: the share of auto repair businesses investing in digital technology rose from 71.1% in 2024 to 75.3% in 2025, up 4.2 points. The figure is consistent across the outlets that reported it, but it remains a regional one: no available source presents it as a national figure, and none of the secondary sources specify the survey's sample size. It stands as a signal of an ongoing trend in one region, not a measurement of Italy as a whole, and it's still a fair example of digital investment in the sector moving in steps rather than sudden technological leaps.

What it means for a repair shop owner

The most immediate competitive edge the data points to isn't buying the most advertised diagnostic tool, but already having, or being able to find, staff with digital skills: that's what over half of the sector's job openings require today. An advanced diagnostic system, however good, is of limited use without someone in the shop who can read and act on its results day to day. It's the same order of priorities that applies in other under-digitized trades: a third of Italian construction companies still don't have a website, and there too the practical basics come before the more advanced tools. It's also the same caution worth applying to any striking figure published by whoever is selling the tool: I wrote about a case in real estate, in artificial intelligence and real estate agencies, where two press releases tied to the same company contradicted each other.

If you're figuring out how to bring your repair shop up to speed with the sector's real demand for digital skills, before chasing the latest announced tool, get in touch: I'll help you work out where to start.