November 20, 2026

Modern data insight and intelligently driven processes can help reduce backlogs in driving

placeholder

Written by Peter Wilson, Director, Industry Principal – Government at Pegasystems 

Between them, the Driver and Vehicle Standards Agency (DVSA) and Driver and Vehicle Licensing Agency (DVLA) have a considerable and somewhat diverse remit and responsibility across the transport sector, encompassing the testing of people operating vehicles, the licensing of them for specific use, regulation of the industry to test road-worthiness, regulation of the industry to train commercial operators, and the type approval and registration of vehicles in both motor trade and individual contexts.

Much has been reported in news outlets over the last 18 months about backlogs across this remit, whether it’s in the DVSA or DVLA space including:

  • Driving test slots for individuals and HGV/PSA testing slots for the transport industry
  • Licence renewals again across individuals and trade “actors”
  • MOT testing backlogs

There are also less well reported knock-on effects across driving instructor capacity, throughout the private and commercial sectors. An obvious response to this is to increase capacity, which usually means humans, whether that’s more instructors, examiners, test facilities and technicians or administrative staff in back-office DVSA functions.

However, as previously observed in similar views on backlogs, just throwing more people at the problem is only part of the answer though and, recruiting, interviewing, and training people is a long-term investment. Organisations must also address the more immediate problem of the bureaucracy that distracts specialist resources from core tasks, impacting their productivity.

To bring that statement to life, the backlog of citizen driving tests is so acute, that it isn’t expected to be cleared until early in 2024 and this has provoked some to call for a temporary model whereby instructors can certify a learner’s attainment of the required driving standard.

Obviously, while understanding the counter argument about conflict of interest and the desirability of a strong independent assessment, it must be possible to use modern data insight and intelligently driven processes to have a more federated model that brings training and accreditation closer? As in other learning environments. Now would be the time for a closely managed trial of that, to test the streamlining of the training to attainment lifecycle.

Thinking about the same problem with an administrative lens, there is clearly more that can be done with data analytics. This includes intelligently associating provisional licence applications with progress on instruction, theory test booking and, ultimately practical test readiness and booking, to prioritise in a meritocratic way when someone books for the end point aka the driving test. It is sobering that there is just a 49.8% passing rate on practical tests according to DVSA’s April 2020 to March 2021 report, which while that may somewhat represent an exacting standard, must still also represent a failure in readiness terms.

Moving onto the trade side of this equation, In October 2021 The Grocer reported that there are “hundreds” of unused HGV test slots every week, despite the backlog for this aspect and against the backdrop of the recent strain on the transport industry with driver shortages. Again, this is somewhat a failure of administration, where a more dynamic approach to align supply to demand would improve the problem.

In terms of driving license renewals, whether citizen or professional, of course there will always be edge cases where there are health check considerations, but given the existence of relatively good online experiences to renew, one has to wonder why the backlogs occur.

At the risk of over trivialising the need for a physical “licence”, fully accepting that the driving licence is often a component in identity in other walks of life, we do need to ask if this is the process aspect that causes delay. If so, why it is, especially in renewals, that we need a physical component over say the way railcards can now be an embedded element in a mobile app.

What all these issues have in common is the human dimension in the form of citizen customers or business actors, and the importance of having more sophisticated understanding about their moments of need and where relevant, their previous interactions and status. Banks call this Client Lifecycle Management, and it is the lifeblood of being able to deal with customers in an insightful, structured, prioritised and, coherent way, regardless of interaction channel or service.

Being able to understand what, when, where and how they interact with your services is not only key to being able to optimise them, but it is also a success factor in doing that across the wider eco-systems that are increasingly found within any large business. In DVSA’s case, think of the driving instruction market, the MOT industry and more.

The DVSA have published the top 10 reasons candidates fail a driving test and, while it is great to publish that, assuming the driving schools are acting accordingly, we have to ask what other insight could be driven out of that information; social and regional demographic view, most and least successful driving schools, to name two.

It is the sign of a healthy evolving “digital first” system to start using data, synthesising insight and looping it back into actionable steps across an eco-system and process framework.

One thing we know for certain is that the government often already has all the data it needs to reduce backlogs, by applying intelligence and automation to address bureaucratic bottlenecks. As a first step to this, it is essential to have a way to break down siloes and create a central hub of information, that can guide the components involved across an eco-system, to improve the way the public service operates day to day.

By bringing data together, applying intelligent process automation and surfacing to humans in the modern digital context expected today, backlogs can be more effectively managed, processing the bulk of simple unambiguous cases, while surfacing complexity and ambiguity in a more efficient way for human review. This fundamentally eliminates the need for valuable human resources to do low-level/low value processing, thus getting a win/win on tackling the problem.