For years, public discussion around autonomous driving has focused on robotaxis. The vision is compelling: summon a driverless car that takes you anywhere, anytime. Yet while this vision continues to capture headlines, a quieter revolution is already taking place.
Autonomous shuttles are moving passengers every day on public roads in multiple countries. The question is no longer whether autonomous transportation works—it is where it creates value first.
From my perspective, having worked with one of Europe’s autonomous shuttle developers, the answer is increasingly clear: commercially viable autonomous shuttle services will scale well before fully autonomous robotaxis become commonplace.
The economics matter more than the technology
Technology is only one part of commercialization.
The real question every mobility operator asks is simple: does autonomy improve the business case?
In many geofenced transport environments—airports, university campuses, hospitals, resorts, business parks and suburban last-mile routes—the single largest operating expense is the human driver.
Unlike robotaxis, autonomous shuttles operate on predefined routes with predictable demand. This creates an opportunity to supervise multiple vehicles from a single remote operations centre instead of assigning one driver to every vehicle.
Even moving from a one-driver-per-vehicle model to one operator supervising three or more autonomous vehicles fundamentally changes the economics of public transportation. As regulations continue evolving, these ratios are expected to improve further.
For operators struggling with labour shortages and rising personnel costs, this is no longer simply a technology discussion—it becomes an operational one.
The last mile is the missing link
Cities have invested heavily in rail systems, metro networks and bus corridors.
Yet millions of passengers still face the same challenge every day: the last kilometre between public transport and their final destination.
This “last mile” often determines whether people choose public transportation or private cars.
Autonomous shuttles can bridge that gap by connecting transport hubs with residential districts, airports, hospitals, industrial sites and commercial centres. They complement existing transport systems rather than competing with them.
This creates benefits beyond convenience.
Better last-mile connectivity increases public transport usage, reduces parking demand and contributes to more efficient urban planning.
Physical AI is becoming reality
Artificial Intelligence is rapidly moving beyond software.
Physical AI—the application of AI in machines operating safely in the real world—is becoming one of the defining technology trends of this decade.
Autonomous vehicles represent one of the most demanding applications of Physical AI because every decision has real-world consequences.
Success requires far more than computer vision.
Vehicles must combine multiple sensors, localisation, mapping, behavioural prediction, safety engineering, cybersecurity and remote supervision into one integrated system capable of operating reliably in dynamic environments.
The challenge is not simply teaching a vehicle to drive.
It is designing an entire operational ecosystem that can be deployed safely and commercially at scale.
Commercial deployment creates better technology
Unlike many emerging technologies, autonomous mobility benefits enormously from real-world operation.
Every kilometre driven generates new operational data.
Every interaction with pedestrians, cyclists, weather conditions and infrastructure improves future software releases.
Companies already operating commercial fleets therefore gain an increasingly valuable competitive advantage over organisations relying primarily on simulation.
Commercial deployment is no longer merely a sales milestone—it has become a core component of product development.
Europe has an opportunity
Europe may not dominate consumer AI platforms, but it has world-class engineering talent, advanced manufacturing capabilities and strong expertise in safety-critical systems.
These strengths align well with autonomous mobility.
If Europe continues supporting deployment alongside research, it has an opportunity to build globally competitive companies in autonomous transportation and Physical AI.
The race will not necessarily be won by those making the boldest predictions.
It will be won by companies delivering safe, reliable and economically sustainable services that customers are willing to pay for.
Looking ahead
The future of mobility is unlikely to arrive all at once.
Instead, autonomy will spread gradually, beginning where the operational case is strongest and where technology already creates measurable value.
That is why airports, campuses, industrial parks, resorts and last-mile public transport are becoming the proving grounds for autonomous transportation.
Robotaxis remain an exciting long-term vision.
Autonomous shuttles, however, are already demonstrating how AI can transform everyday mobility—not as a laboratory experiment, but as a commercial service transporting real passengers today.
For the logistics and mobility industries, the most important question is no longer whether autonomous transportation will become mainstream.
It is where commercial adoption will happen first.
The evidence increasingly points toward autonomous shuttles.
