AI for intelligent inspections at DVSA

Every year in the UK over 40 million vehicles receive an MOT standards test. This is big business - generating £1 billion for more than 80,000 testers in 23,000 garages across the country. However, in an industry this size, it is inevitable that some vehicles are not tested properly - with lethal implications. Mitigation of this through the inspection of authorised garages was resource intensive as there was limited knowledge to target inspections to best effect.

DVSA’s digital team wanted to explore data-driven approaches to help them conduct intelligent inspections of authorised garages carrying out MOTs, to ensure that vehicle standards are enforced. They developed an approach that analyses this vast amount of testing data and uses AI to generate a real-time risk profile.

Initially the DVSA lacked a precise view of what constituted a high-risk tester profile.  Turns out that when looking for a needle in a haystack, it pays to observe the distribution of other needles. The agency first identified outliers in the data by using a density-based anomaly detection algorithm.  This grouped testers with similar profile scores and identified the outliers as potentially high-risk testers.

The results focused on individual MOT tester’s performance, took into account changes to testing behaviour, refreshed regularly to ensure current risk scoring, and included test volumes, frequency, duration and pass rates, as well as disciplinary history. This score combined with the outcome of any previous enforcement action to produce the risk rating.

Since DVSA introduced this approach to risk rating months ago, the new system has found more cases than we would have expected to see using the previous approach to investigation. This has highlighted some garages that we have been able to stop testing, immediately making the roads safer.

The AI flagged a garage owner in Devon, and on further investigation the DVSA found that the garage had carried out more than 300 fraudulent MOTs. This led to the garage owner being prosecuted, given a suspended prison sentence and banned from testing.

The image used for this article is licensed under the Open Government Licence v3.0.