How Autonomous Vehicle Operators and Councils Could Help Each Other Out
AV operators need real-world edge cases. Councils need capacity gains. Early partnerships could benefit both.
Behind every bin collection, gritted road, and pothole repair there’s a council vehicle fleet that most residents rarely think about. UK local authorities operate thousands of vehicles: refuse trucks, gritters, street sweepers, pool cars for staff, minibuses for school transport, and unmarked cars for housing officers and building inspectors. These fleets are not glamorous, but they quietly determine the cost, quality, and availability of services that affect daily life.
Councils face shrinking budgets alongside growing service demands. Meanwhile, autonomous vehicle technology is maturing, but even after 100 million miles, operators like Waymo still encounter edge cases: vehicles not stopping correctly near school buses, no robust procedures for power outages. Delivering any service, autonomous or not, means handling the unexpected.
This creates an opportunity for mutual benefit. Autonomous vehicle (AV) operators need controlled environments to refine their technology and learn edge cases. Councils need better information on where services are needed and the capacity and productivity gains to deliver them at cost. Early partnerships give operators real-world deployment experience while expanding what councils can deliver, whether that’s supporting school transport, or autonomous vehicles reporting potholes as they drive.
But where to start? Not every council service suits early AV deployment, and not every operator is ready for the messiness of public service delivery.
What Makes a Strong Council AV Use Case?
The clearest opportunities are where driving is the constraint, services limited because councils can’t afford enough drivers, can’t staff antisocial hours, or where a staff member, such as a social worker, inspector spends significant time travelling between tasks during their work day. Nevertheless, each deployment builds shared knowledge that neither side has alone. Operators discover edge cases that don’t show up even after a million miles of simulation. What happens if a an autonomous street sweeper finds a fox den, how should it proceed? Meanwhile councils would learn the operational processes of autonomy: where charging infrastructure goes, what the maintenance cycle looks like, who to call when something breaks. That groundwork makes later, harder and more ambitious deployments possible.
With that lens, here are the most promising applications, these are ordered by how soon they could realistically happen, starting with services that could use robotaxis as they already exist, then moving to operations that will require purpose-built autonomous vehicles.
Productivity Gains: Working on the Move
The most useful service efficiency improvement may not be from the capabilities of vehicles but from saving staff time. Many council officers spend their work-days travelling between site visits. Building control inspectors, housing officers, environmental health officers, social workers: the pattern is similar across roles. Drive to a location, conduct an inspection or check-in, write up notes or a report, then drive to the next appointment. Repeat.
Currently, driving time is dead time. Officers face a constant trade-off: more visits means less time for the paperwork and follow-up that each visit generates. Workloads are shaped around what can be achieved between journeys, not around what the service actually requires.
If the vehicle drives itself and is economical for a council to operate, that time opens up. An inspector reviews plans for the next site while in transit, or dictates a report from the visit just completed. Calls that currently wait until the officer is back at a desk happen between appointments. The result is either more visits per day, or the same number of visits with better quality follow-through, or some combination of both.
The technology exists. The question is whether it becomes cheap enough to outcompete contracted taxis and pool cars. Robotaxi services are launching in London, and as more competitors enter, prices will fall. Councils will find it progressively easier to shift journeys onto autonomous options: first a few staff trips, then more.The fastest route of adoption would be operator-council partnerships offering discounted rates to public sector clients, building goodwill while demonstrating value.
This raises practical questions. Who gets access to pooled autonomous vehicles first? Should social workers be prioritised over environmental officers? How do you measure staff efficiency gains to know the investment is worth it? Will staff be willing to use ‘transport time’ to do work? Each authority will exercise its own judgement, but the useful framing for councils not just “how can autonomy work for us?” but “how does our work change to make the most out of autonomy?.”
School Transport
Councils have a statutory duty to arrange home-to-school transport for eligible pupils, particularly children with special educational needs and disabilities (SEND). For 2025/26, that obligation will cost nearly £2 billion nationally, triple what it cost a decade ago. In practice, it means a patchwork of council-operated minibuses and contracted private hire vehicles, often individual taxis at £8,900 per SEND child per year and £3,100 for non-SEND child transport.
Already we see parental demand for autonomous school transport services. In San Francisco, parents have been sending their children to school in Waymo robotaxis. Waymo launched a teen service allowing 14-17 year olds to book rides independently. This gives a glimpse into parental trust in using this technology but a similar fast-take up will not happen spontaneously by British parents as most children either walk or use public transport to get to school. There is no mass-scale autonomous school bus service yet, but as costs fall and more efficient shuttle designs become common, this will change.
For SEND transport specifically, autonomy widens the pool of passenger assistants. Currently these roles often require a driving licence even when the core skill is caring for vulnerable children. Removing the driving requirement opens the job to more people, including social workers and carers who do not drive.
Councils will need to be prudent. Understanding the true cost of delivery matters, because any future operator may seek to undercut current contracts only slightly rather than pass on the full savings. Authorities that lock themselves into expensive long-term agreements now may find themselves paying well above market rates within a few years.
Monitoring and Enforcement and Multi-Purpose Vehicles
This is the nearest-term opportunity for autonomous vehicles in council operations, and it requires no new vehicles at all. Robotaxi operators already collect continuous road-condition data from their fleets. Data-sharing agreements with local authorities turn every autonomous taxi into a passive inspector, flagging potholes, damaged signs, blocked gullies, overhanging branches, and failed streetlights as they drive.
The same logic extends to enforcement. Cameras that detect road defects can spot dumped mattresses, overflowing bins, or vehicles parked illegally. Over time, this can lead to multipurpose vehicles. An autonomous street sweeper working at 4am simultaneously reports fly-tipping or major traffic violations as they happen.
This shifts how councils think about fleets. Rather than buying vehicles for single functions, the question becomes: how many tasks can each vehicle usefully perform across a 24-hour cycle? A vehicle that sweeps at night, monitors road conditions at dawn, and patrols for fly-tipping during the day starts to earn its keep. The same sensors serve multiple purposes. The same miles generate multiple outputs.
For councils, the practical question is how to receive and act on this information. More data is only useful if there are processes to triage and act upon it. A flood of pothole reports is a problem if the repairs team cannot keep pace. But that is an operational challenge, not a technological one, and councils that begin thinking through these workflows now will be better placed when the data starts arriving.
Pothole Prevention
Monitoring potholes is the first step, but the technology is slowly moving towards active repair. Highway Safety Inspectors currently drive designated routes on set schedules, visually assessing road conditions and logging defects. Main roads might be inspected monthly; minor roads less frequently. Between inspections, potholes form and worsen. Councils often learn about problems only when residents complain, by which point a crack has become a hole requiring a full repair team. Preventative maintenance cut down on costs massively.
Autonomous vehicles can close this gap. The UK startup Robotiz3d has developed a vehicle that detects road cracks and seals them before they become potholes, claiming 90% cost savings and 70% faster repair times compared to traditional methods. The firm is at early stages, but the principle is clear: continuous scanning plus immediate treatment means fewer potholes forming in the first place.
Instead of dispatching crews to fix what residents have reported, the work for councils becomes deciding which roads to scan and treat most frequently. Local knowledge on the importance of particular roads is crucial here, the job becomes a more pleasant one: less firefighting, more prioritisation.
Street Sweeping
Autonomous sweeping can happen at 4am on Sunday morning, after Saturday night’s crowds have dispersed. Additionally, autonomy eases post-event sweeping operations: after a football match, market day, or fireworks display. Longer term, adaptive scheduling based on sensor-detected debris can replace fixed cycles altogether with continuous autonomous cleaning services.
In terms of existing deployments, the Chinese autonomous vehicle company WeRide has developed the ‘Robosweeper’ an Autonomous Sanitation Vehicle with brushing and hosing capabilities deployed in 10+ cities internationally but currently the vehicles service capacity per charge is small, at about 12 hectares or 0.12 km².
If post-event cleaning becomes cheap and automatic, one friction point for permitting events disappears. Councils often require organisers to submit waste management plans or pay cleaning deposits. This is rarely the decisive factor in whether an event goes ahead, but it adds administrative burden, particularly for smaller community events. Autonomous street sanitation does not solve the harder constraints like noise and traffic disruption, but it mitigates one item from the list.
Conclusion
The opportunity here is mutual. Councils hold something AV operators need: real-world environments full of edge cases, institutional knowledge about service delivery, and the political legitimacy to deploy autonomous vehicles in public spaces. Operators hold something councils need: the technology to break the link between staffing constraints and service capacity. Neither side can realise the full value alone.
The deeper shift is in how councils think about their operations. Today, service design starts from what can be staffed. A sweeper runs at 10am because that’s when drivers are available, not because that’s when sweeping works best. School transport costs what it does because the market for drivers sets the floor. Autonomy breaks that link. Services can be designed around what is needed, not what can be staffed. New roles in fleet management, data triage, and autonomous vehicle maintenance would emerge. The overall effect is more services delivered per staff member, not fewer staff. At first Operators will need to meet councils halfway. Data-sharing agreements and genuine engagement with the messiness of public service delivery will do more to build the market than waiting for councils to issue tenders.
Councils that recognise what they bring to the table can negotiate as partners, not just purchasers. The work required is mutual. So is the reward.
