Autonomous depot operations recover 25 to 35 minutes per bus per day from non-revenue movements. Most business cases value that time at the driver's hourly cost, around €35–45. That is too low. When an operator cannot hire enough drivers, every hour spent moving buses around the depot is lost revenue, because the same hour could have been spent carrying passengers.
We published the standard case ourselves
A year ago we published an ROI framework for depot autonomy. The logic was simple: measure the time drivers spend on parking, charging approach, wash bay positioning and dispatch staging, multiply by the full hourly cost of a driver, and compare against the investment.
For a 100-bus depot at 25 minutes per vehicle per day, that is roughly 42 driver-hours a day and €500,000 to €700,000 a year. MZA Warsaw ran the same calculation per vehicle. At up to 35 minutes per bus, it came to over €88,000 across a 12-year lifecycle, which is around €730,000 a year for a depot of 100 buses.
Those numbers hold. There is a second part of the case that the model does not capture.
The standard model assumes you cut hours
If you value a recovered hour at the driver's wage, you are assuming the operator ends up paying for fewer hours. Fewer hours worked, smaller wage bill, money saved.
European operators are trying to hire more drivers. Most cannot fill the positions they already have. According to the International Road Transport Union, around 105,000 bus and coach driver positions are unfilled across Europe. The gap has grown by 54% since 2022. The average driver is over 50, and only 3% are under 25. Retirements alone could push the shortage toward 275,000 by 2028. Operators have raised wages, improved conditions and run recruitment campaigns. The gap is still there.
So the hour does not leave your wage bill. You pay the driver either way. What changes is what the driver does with that hour, and that is what decides its value.
There are two cases:
You are fully staffed. The hour comes off overtime, or off an agency invoice. It is worth roughly the wage, and the standard model is right.
You are short of drivers. The hour becomes service you could not otherwise have run. So every hour a driver spends moving buses inside the depot is revenue the operator does not earn. This is lost revenue, and it is a bigger number than the wage saving.
Most operators we speak to are in the second case.
The value depends on the driver staying
The value comes from the driver staying and moving to the line. If the recovered time were simply taken out of the wage bill, it would be worth the wage and nothing more. It is worth more than that because the same driver now spends those minutes carrying passengers.
There is a second effect. The time we recover is the worst part of the shift: 23:00, tight lanes, reversing, fatigue, someone from the depot office asking you to move the bus off the wash lane before you can go home. Those are the hours drivers most want back. Better end-of-shift conditions help operators keep drivers, and keeping drivers is itself a way to close the gap. Both effects work in the same direction.
How much revenue is being lost
This depends on your contract. Here is how to work it out for yours.
Gross-cost contract. You are paid for each vehicle-hour or vehicle-kilometre you deliver. The driver is roughly half the operating cost of a service hour. The rest is vehicle, energy, maintenance, depot and overhead. So the payment for delivering that hour is higher than the driver's cost, in most cases by around a factor of two. The extra cost of running the hour is also lower than usual, because you are already paying the driver. Check the ratio against your own contract rate.
Net-cost contract. The hour is worth the fare revenue on the trip, plus the passengers you keep. A cancelled bus loses the fare today, and over time it loses the passenger who decides the route is unreliable.
In-house municipal operator. There is no external payment. The hour is worth the overtime premium you did not pay, the agency rate you did not book, and the timetable you did not cut. Ask your operations director what a cut service costs. That cost rarely appears in a depot automation business case, even though depot automation is one of the things that prevents it.
One more cost usually sits outside the model: penalties for service you did not run. Lost-kilometre rates are contractual, measured and enforced. Ireland's Department of Transport told the Public Accounts Committee that reliability is measured using a lost kilometre rate. This is the share of services that failed to operate for reasons inside the operator's control, and driver shortages were named among the causes. Penalties across public transport operators there came to more than €17 million.
Turning hours into drivers
The shortage is counted in people, so convert the hours into people.
A 100-bus depot at 25 minutes per vehicle per day recovers about 15,200 driver-hours a year. At 35 minutes it is about 21,300. Assume about 1,700 working hours per driver per year, after holidays, sick leave and training. Assume you recover 75% of the theoretical time, because the rest goes to handover at the gate and the arrival checklist.
That is the equivalent of seven to nine full-time drivers per depot — drivers you already employ, returned to service.
You get them without a recruitment campaign, without paying for licences, without waiting three to six months for a new driver to become productive, and without raising wages to compete with the operator in the next city. In our conversations with operators, this number gets more attention than the cost saving.
From here, three steps
Work out the number for your depot. Decide whether it is worth doing. Then set the requirements for what you buy.
Step 1: Work out the number
Build it from five of your own figures.
Step 2: Decide whether it is worth doing
Set the annual value against the annual cost, and against what you are already spending on the same problem.
The cost side has four parts: a one-time cost per bus for the autonomy system, a one-time cost per depot for mapping, Mission Control installation and training, a recurring support cost per bus-year after warranty, and your own internal effort on training, procedures and integration. The per-depot part is around €25,000 and it spreads across the fleet: about €1,250 per bus across 20 buses, about €250 per bus across 100.
Then compare it with the alternatives you already pay for. Recruitment campaigns. Licence funding. Agency drivers. Overtime premiums. Wage increases to compete for the same small pool. Reduced timetables. These are recurring costs, and none of them adds capacity that stays.
Two things make the decision easier to take. Grants and co-funding can cover part of the capital cost while the operational value stays with you. And the commitment can be staged: one bus to confirm feasibility in your depot, ten to confirm it scales, and a larger number at fleet renewal. The first step is small enough to test the assumptions before committing budget.
The decision also has a date attached. If a fleet renewal falls in the next 24 months, that is the point where autonomy is cheapest to introduce, and the decision needs to be made before the specification is written.
Step 3: Know what you are buying
Once the decision is made, seven areas determine what you actually get. Ask suppliers about them, and ask for the answers in writing.
Work out your own number
The figures in this article are averages across European depots. Yours will be different, and the difference matters.
The Smartbus Onboard Program is a site-specific assessment of your depot: measured shunting time, a quantified business case, a depot layout review and a fleet compatibility check. It gives you the number to take into Step 2.
→ Start with the ROI calculator to test your assumptions in a few minutes, or join Smartbus Onboard Program to arrange your depot assessment.