impair.ai
Impair.ai was a machine learning concept for predicting the likelihood of future issues with vehicles. More broadly, it is an example of how artificial intelligence can be applied to vehicle history data to estimate how reliable a particular car is likely to be.
How AI can predict car problems
The UK is unusually rich in public vehicle data. The DVSA publishes the MOT result of every test carried out since 2005, including failures, advisories, defect categories and recorded mileages, and the DVLA holds registration details for every vehicle on the road. Machine learning models can be trained on millions of these records to spot patterns – for example, that a certain engine and model year tends to develop suspension or emissions problems at a particular age or mileage.
Given a specific car’s registration, such a model can compare its history against thousands of similar vehicles and estimate the probability of it failing its next MOT, or of certain components needing attention soon. Signals like recurring advisories (corrosion, brake wear), inconsistent mileage readings, or gaps in the test history all feed into that picture.
What predictions can and cannot tell you
These predictions are probabilities, not guarantees. A car flagged as higher risk may run faultlessly for years, and a low-risk car can still break down – condition depends heavily on how an individual vehicle has been driven and maintained. Treat AI reliability scores as one input alongside a physical inspection, a test drive and the documented service history, not as a substitute for them.
To see the factual record any prediction would be built on – MOT results, mileage history and more – start with a car history check.