Key Points
- Artificial intelligence cameras flagged nearly 19,000 possible mobile phone and seatbelt offences during a two-week West Midlands trial.
- The cameras operated outside Ladywood Fire Station in Birmingham and Fallings Park Fire Station in Wolverhampton.
- The published site figures differ between reports, with totals ranging from 18,995 to 19,145 potential offences.
- No fines or penalty points were issued because the exercise assessed the technology and scale of possible offending.
- Authorities are evaluating whether the cameras could support future police enforcement.
Inverted pyramid:
Birmingham (Glasgow Express) November 18, 2026 – Artificial intelligence cameras flagged almost 19,000 potential seatbelt and mobile phone offences during a two-week road safety trial in the West Midlands. The system analysed vehicles outside Ladywood Fire Station in Birmingham and Fallings Park Fire Station in Wolverhampton, identifying drivers who appeared not to be wearing seatbelts or who appeared to be using mobile phones. No enforcement action followed, and the detections were not confirmed offences or penalties.
What did the West Midlands camera trial find?
The RAC reported that Ladywood produced initial scans of 11,144 possible seatbelt offences and 1,163 possible mobile phone offences. At Fallings Park, the corresponding figures were 6,081 and 757. Those figures total 19,145, although other reports citing West Midlands Combined Authority figures gave lower counts of 11,045 seatbelt and 1,163 phone detections at Ladywood, plus 6,030 seatbelt and 757 phone detections at Fallings Park, producing 18,995 potential offences.
The RAC said the cameras were supplied by AECOM and Acusensus. Human operators checked the technology using a sample of 8,024 images, according to reports citing the West Midlands Combined Authority.
Why were no drivers prosecuted?
The trial was designed to assess the possible scale of unsafe driving and test the accuracy and usefulness of the camera system. The reported detections were initial artificial intelligence flags, rather than proven offences. No fines, penalty points or prosecutions resulted from the exercise.
West Midlands Police and Crime Commissioner Simon Foster said the results were being evaluated to determine how the technology might be used in future police enforcement. West Midlands road safety commissioner Mat MacDonald said the volume of detections demonstrated the potential scale of mobile phone use and seatbelt non-compliance.
What could happen to drivers in future?
Government guidance states that drivers must not hold and use a mobile phone while driving, including when stopped at traffic lights or queuing, subject to limited emergency exceptions. Holding and using a phone can result in six penalty points and a £200 fine. Failing to wear a required seatbelt can lead to a fine of up to £500, while failing to restrain a child properly can carry a fine of up to £500.
What is the background to the development?
The West Midlands Combined Authority funded the trial with police and fire-service partners, placing the cameras at two fire stations during March. The RAC reported that a separate Devon and Cornwall trial began in 2023 on the A30 and A38 and was later extended to the M5, with the programme due to continue until April 2027. The RAC also said a similar programme involved 10 English police forces in 2024. The combined authority linked 2025 road deaths on United Kingdom roads to mobile phone use and unbuckled seatbelts, citing 32 and 133 deaths respectively.
How could the findings affect drivers and passengers?
If authorities adopt the technology for enforcement, drivers who use handheld phones or fail to wear seatbelts could face a greater likelihood of being detected, particularly on busy urban roads. The findings may also affect passengers and families because drivers are responsible for ensuring children use the correct restraint. Any future enforcement would require authorities to establish operational arrangements and confirm suspected offences through the relevant legal process.
