United KingdomPublic sector2025 – ongoingDriver and Vehicle Standards Agency
Machine-learning risk ratings for MOT inspection priorities
Approximately 150 DVSA officers used machine-learning risk ratings daily to prioritize MOT garage visits
Problem
DVSA previously prioritized garage visits by the time since the last visit. It needed to direct inspection resources toward testers and garages requiring attention.
Approach
Kainos developed the initial tool before handing development ownership to DVSA. A local outlier factor model used MOT testing data to generate risk ratings, refreshed monthly. Officers considered these ratings alongside other information when selecting visits.
Results
Takeaway
The model helped select inspection targets. Disciplinary decisions required evidence from subsequent investigative activity.
Categories
- Size
- Large enterprise
- Industry
- Public sector
- Business function
- Production and quality
- Innovation type
- Process innovation
- AI technology
- Predictive analytics and ML
- Adoption
- Partner co-development
- Evidence
- Company's own claim
Sources
Source summaries
DVSA documented a production model that prioritized MOT inspection visits and was used daily by approximately 150 officers. It described human review but did not quantify improvements in inspection effectiveness.
Full details are in the original sources.