Europe is not pursuing autonomous trucks simply because driverless vehicles look futuristic. The immediate problem is freight capacity. The latest IRU data indicate that around 502,000 professional driving positions are unfilled across Europe. Sixty-five per cent of operators identify the shortage as their main business concern, while roughly two-thirds have rejected new contracts because they could not find enough drivers. By 2030, about 660,500 European truck drivers are expected to retire.
Autonomous road freight should therefore be viewed as a way to separate the hardest kilometres to staff from the tasks that still require people. Long, repetitive motorway sections are the strongest candidate. Customer communication, load securing, paperwork, difficult yard manoeuvres, urban access and first- and last-mile delivery remain much less suitable for removing the driver.
What an Autonomous Truck Actually Is
The term autonomous truck is often used too loosely. A vehicle that can keep its lane and distance is not equivalent to one that can operate without a person in the cab.
- Level 2 combines steering and speed control, but the driver continuously supervises the road and remains responsible.
- Level 3 drives under limited conditions but may require the driver to take over.
- Level 4 operates without a driver only inside a defined Operational Design Domain, such as particular roads, speeds, weather limits and terminal procedures.
- Level 5 means unrestricted automation in all places and conditions. It is not a realistic near-term target for European freight.
Only driverless Level 4 operation can directly reduce the human driving hours required. Driver-supervised automation may improve comfort, consistency, fuel use and safety, but it does not eliminate the position. Current European pilots demonstrate this distinction: some carry real freight while still requiring a safety driver.
Why Hub-to-Hub Transport Comes First
The most credible early model is hub-to-hub transport. A human driver collects the trailer and brings it to a controlled site near a motorway. A Level 4 tractor completes the long motorway section. At the destination hub, another driver handles the complex final kilometres.
This arrangement also addresses European working-time limits. A driver may normally drive nine hours per day, extend this to ten hours twice per week, drive no more than 56 hours per week and must take a 45-minute break after 4.5 hours. A driverless motorway section could reduce dependence on those limits, although loading, unloading, charging or refuelling, inspections and terminal delays would remain.
What Europe Is Actually Testing
A motorway demonstration, a supervised commercial journey and a driverless operation in a closed site should not be presented as the same level of maturity.
- MAN tested an ATLAS-L4 tractor on approximately ten kilometres of Germany’s A9 motorway in April 2024. The project focused on a production-oriented Level 4 vehicle for hub-to-hub freight.
- Scania has operated a roughly 300-kilometre freight route between Södertälje and Jönköping. A safety driver remains in the cab, and the first and last sections are driven manually.
- An IVECO S-Way with PlusDrive has been tested on a real German route. It automates several motorway functions but remains supervised by a driver.
- Seven autonomous Volvo FH trucks transport limestone over a five-kilometre Norwegian mining route with tunnels and inclines. Safety drivers were removed in 2023, and more than one million tonnes have been transported. The customer buys transport capacity rather than only vehicles.
- In 2025, a cabless heavy vehicle completed a controlled public-road demonstration in Belgium. A separate MODI demonstration crossed the Sweden–Norway border without a person on board. These were proofs of capability, not unrestricted commercial operations.
- At a container terminal near Tallinn, driverless terminal tractors are being introduced without safety drivers. Restricted access, short routes and low speeds make this easier to control than international motorway transport.
Why Mines, Ports and Terminals Lead
A Level 4 system operates only inside its Operational Design Domain. This may define the route, road type, speed, visibility, temperature, rainfall, lane markings and situations in which the vehicle must stop.
A closed terminal offers low speed, known routes and restricted access. A motorway corridor is faster but has relatively predictable geometry and limited manoeuvring. Urban delivery combines pedestrians, cyclists, parked vehicles, roadworks and frequent exceptions. Success at 20 kilometres per hour in a mine therefore does not prove safe performance at motorway speed.
The MODI project uses the 1,200-kilometre Rotterdam–Oslo corridor to study border crossings, ports, terminals, tolling, customs, connectivity and transitions between controlled sites and public roads across four national borders.
A Practical Route Scorecard
A carrier or shipper can use this pre-feasibility screen. Give each item zero points for unfavourable, one for manageable and two for favourable conditions.
- Share of the route on motorways
- Consistency of the route and departure times
- Stable daily freight volume
- Availability of return loads
- Controlled hubs at both ends
- Number of national borders and permit regimes
- Exposure to mountains, severe winter weather and frequent diversions
- Standardisation of trailers and loading procedures
- Reliable connectivity and technical support
- Backup capacity when automation is suspended
As a working rule rather than a regulatory standard, 16 to 20 points indicate a strong early candidate, 11 to 15 suggest a possible pilot, and below 11 usually favours conventional transport. A regular 400- to 700-kilometre lane with more than 80 per cent motorway driving, several daily departures, standard semi-trailers, return freight and controlled yards is far more promising than multi-drop work with irregular addresses, oversized cargo or frequent deviations.
The Economics of a 600-Kilometre Lane
The business case is not simply the driver’s salary removed from the cost sheet. Annual value includes avoided driving labour on the automated section, additional productive kilometres, better tractor utilisation and possible reductions in energy use and damage. Costs include the autonomous system, software, remote supervision, sensor calibration, hub transfers, communications, cybersecurity, reserve capacity, downtime and maintenance.
Three scenarios show why route design matters:
- Driver-supervised automation keeps the driver in the cab. It may reduce fatigue and improve consistency, but offers little direct labour saving.
- Level 4 with weak utilisation can fail economically if the truck returns empty, waits for hours at each hub or requires one remote operator per vehicle.
- Level 4 with high utilisation is stronger when the truck repeats a motorway lane, carries freight both ways, has short terminal dwell times and uses safe one-to-many remote oversight.
One middle-mile study estimated total cost reductions of approximately 15 to 20 per cent for a specific autonomous heavy-truck model. That result should not be transferred mechanically to Europe because wages, road charges, insurance, hub costs, regulation and annual loaded kilometres vary widely.
The 24-Hour Truck Myth
Automation removes the need for a driver to rest; it does not remove the rest of the logistics process. A truck may still wait for a free dock, completed documents, a checked trailer, fuel or charging, sensor cleaning, technical inspection, better weather or remote assistance.
A vehicle capable of moving for 20 hours per day may still be uneconomic if it spends five hours waiting across two terminals or returns empty. The useful question is how many hours the complete operation can provide freight, a ready trailer, an available dock and an open route.
This creates demand for transfer hubs with trailer storage, maintained autonomous lanes, automated entry, tyre and light inspection, load and coupling checks, sensor cleaning, safe-stop areas, reliable communications, recovery staff and integration with transport and warehouse systems.
Remote Operations Are Not Office-Based Driving
Remote operations usually mean monitoring fleet status, confirming unusual situations, coordinating with police or terminals, arranging roadside assistance and organising physical recovery. Limited remote manoeuvring may be possible where law and system design allow it, but the objective is not to have a person continuously drive every truck from a screen.
The critical question is how many vehicles one operator can supervise safely. One employee per truck removes much of the labour advantage. One-to-many supervision requires evidence that simultaneous incidents can be handled without delaying a safe response.
Europe’s Regulatory Reality
EU Regulation 2022/1426 provides a technical type-approval framework for automated driving systems, including specified hub-to-hub applications. Technical approval is not permission to run without a driver on every road. National rules still govern route authorisation, remote oversight and incident responsibility.
The European Commission’s March 2025 automotive action plan acknowledged that testing and deployment rules remain fragmented. It proposed cross-border testbeds, regulatory sandboxes, automated-driving corridors and further expansion of the hub-to-hub framework.
In June 2026, UNECE adopted the first global regulatory framework for fully autonomous driving systems. It establishes common expectations for safety management, credible testing, safety-case validation and continuous in-service monitoring, but it does not automatically authorise cross-border operation.
Safety, Software and Data Must Be Measurable
Heavy goods vehicle crashes accounted for 13.8 per cent of EU road deaths in 2023, or 2,829 fatalities. Most victims were other road users rather than truck occupants. Safety performance is therefore a public requirement, not merely a selling point.
Before committing to a pilot or contract, an operator should request:
- Kilometres completed in the same Operational Design Domain
- System interventions per 1,000 kilometres and their causes
- Results in rain, snow, fog, glare and poor markings
- Behaviour after tyre failure, fallen cargo or sudden obstruction
- Redundancy in braking, steering, power and communications
- The procedure for reaching a minimal-risk condition
- Independent incident investigation and access to vehicle data
- Rules for updates, map errors, cyberattacks, data ownership, support termination and independent servicing
UN Regulation 155 requires a cybersecurity management system, while UN Regulation 156 covers software-update management. They support continuous risk control and should not be treated as one-time certificates that remove operational responsibility.
The Most Likely European Future
Autonomous trucks are unlikely to replace professional drivers across Europe. Their first meaningful role will be narrower: moving standard trailers over repeated motorway corridors, inside mines and terminals, and between carefully prepared hubs.
The labour effect will probably be a redistribution rather than simple elimination. Fewer people may spend entire shifts on repetitive motorway driving, while more work shifts toward first- and last-mile transport, yard operations, maintenance, remote supervision, exception handling and customer-facing tasks.
The technology can help protect freight capacity as Europe’s driver workforce ages, but only when the route, hubs, regulation, safety evidence and economics are designed as one system. A driverless tractor on its own is not a freight solution. A reliable operating model around it is.