For more than a decade, the autonomous bus has been "almost ready."
The first demonstrations with safety drivers ran around 2015. Ten years later, that is still, broadly, what we see: a vehicle with a person sitting behind the wheel, ready to take over. The technology has improved enormously. The business model has not moved at all.
The reason isn't a lack of ambition. It's physics, safety and economics — in that order.
A city bus is a heavy vehicle carrying dozens of standing passengers through a chaotic urban environment. The hardest single problem in that setting is the one engineers call the minimal-risk manoeuvre — in plain language, what the vehicle does when something goes wrong. At 50 km/h, with people standing, physics doesn't let a bus stop instantly, and an emergency stop is often the last thing you'd want. The system has to be intelligent and reliable enough to handle the emergency itself.
So far, the industry's answer has been to put that responsibility back onto a human: the safety driver. From a safety point of view, sensible. From a business point of view, it quietly dismantles the entire case. You've replaced a driver with a safety driver — and you're still paying for all the technology on top. That isn't an autonomous bus. It's a more expensive bus.
It shows in the numbers. I can point you to plenty of companies that have lost tens of millions — sometimes billions — chasing autonomous vehicles. I can't point you to one that has made meaningful money from an autonomous bus.
Now change two variables. Take the speed down to 10–15 km/h. Take the passengers out. Put a fence around the operating area.
Everything changes.
At low speed, in an empty vehicle, the system has time to react with margin. If something unexpected appears, it slows or stops — and stopping introduces no risk, because there's no one standing on board and nothing travelling fast. The scenarios that are critical on a public road become manageable, low-risk events inside a controlled yard. The depot turns the hardest version of the problem into a solvable one.
That's not a downgrade of the autonomy ambition. It's the place where the ambition becomes deployable — and pays for itself — today.
This is what we call a smartbus: an ordinary city bus that can move itself around the depot, without a driver, inside a private and controlled environment. Same homologated bus on the road; autonomous mode active only inside the fence. The road-going vehicle doesn't change. The economics of the depot do.
smartbus = ordinary bus + SmartDEPOT™.
The autonomous era for public transport doesn't begin on a downtown boulevard with a safety driver hovering over the wheel. It begins behind the fence, at 15 km/h, with a real return on the investment. It begins in the depot.
Want the full case? We put the whole argument — the business case, the safety framework, the maturity model and case studies from operators across Europe — into one report: The State of Smart Depots 2026.
The future does not ride an ordinary bus.