DVSA + tacho + WTD + OCRS. One substrate.

AI can flag hours-of-service (HOS) violations from tachograph data, but it cannot replace the legal obligations under EU 2016/799: driver-card and vehicle-unit data must still be downloaded (cards every 28 days, vehicle units periodically) and retained for at least 1 year. AI is best used to detect breaches early — overrun driving, missing rest, manipulated cards — so the transport manager can act before a DVSA roadside check. MEOK's tacho-audit MCP produces the signed anomaly evidence.

The road transport WTD (2002/15/EC) caps driving and working time: max 9 hours daily driving (extendable to 10 twice a week), 56 hours weekly driving, 90 hours over a fortnight, and a minimum 45-hour weekly rest. These sit alongside EU 561/2006 drivers' hours. AI shift-scheduling must respect all of these limits simultaneously, which is where most manual rotas fail — our scheduling check validates a roster against WTD and 561/2006 before it is published.

DVSA's Operator Compliance Risk Score (OCRS) and Earned Recognition scheme judge operators on maintenance and traffic compliance history. Earned Recognition requires monthly KPI uploads and demonstrable systems. AI anomaly detection helps keep your green OCRS band by catching missed downloads, overdue maintenance, and HOS infringements before they hit your score — and the audit trail is exactly what the DVSA auditor expects to see.

The DVSA Operator Licence (Standard National or Standard International) is judged on financial standing, professional competence (a CPC holder), and good repute — not on whether you use AI. But the systems you rely on for compliance must be robust and auditable. If AI underpins your drivers' hours or maintenance compliance, you should be able to show the Traffic Commissioner that the system is validated and that a human transport manager retains effective control.