The rise of the Autonomous Bus: what do they have to offer?

The media has been quite exercised lately by the planned trial of Autonomous Buses across the Forth between Ferrytoll and Edinburgh Park.

In early May, a member of EBUG’s Committee took part in a workshop run by Edinburgh Napier University as part of the ‘CAVForth’ trail.

The day’s agenda included: Introduction to the research, project overview, pre-trial surveys, benefits and how to realise them, risks and mitigation, drivers and barriers (to name a few)…

It did not include any trial runs of the buses. On-ground trials, with passengers, start later in 2022 (Monday – Sunday, up to every 20mins; probable route number AB1).

The introduction revealed the answer to ‘what are ABs for?’ (reframed as ‘what are their expected benefits?’) is:

  • Safety: elimination of human error
  • Reliability: ABs could
    1. potentially overcome unexpected staff unavailability
    2. continuously adjust optimal driving to meet local circumstances
  • Reduce fuel use

However, research so far indicates the main benefit may be reducing driver workload, allowing staff to redeploy en route to focus on passengers’ personal security. The research also shows that full automation is not acceptable to passengers or regulatory authorities. Therefore, there would always be a member of staff on board; in a different role to now, but present.

But here’s an unresolved paradox. If onboard staff monitor, but don’t engage with, the driving function, their attention deteriorates after 20 minutes; whereas an active human driver is considered able to focus through a shift of several hours. During the CAVForth trial, two staff will be on-board: a driver/monitor and a ‘bus captain’. We’ll call the bus captain the ‘conductor’, which better reflects the role.

Clearly, two-person bus operation is financially unviable beyond an experimental phase. As a no-person operation is impractical, this suggests not cutting staff headcount, but significant retraining/reorientation of job descriptions. Probably with cost implications.

As for safety, the potential for eliminating human error is understood. However, the incidence of personal injury or death on a bus is already 60% lower than in cars/vans, and is significantly skewed towards the ‘minor injury’ end of the scale. So it is questionable whether this is a useful return.

Reliability is linked to the current high vacancy rate bus operators face. It was suggested that automatic bus operation could overcome situations where a driver is unavailable. But, as discussed above, a staff member is always required.

It is easy to see how a computer ‘brain’, continuously gathering data on traffic conditions, could drive optimally to those conditions. But buses use fixed routes, except (possibly) where there are long gaps between bus stops. As buses are primarily delayed by congestion on those fixed routes, the benefit may be marginal.

The CAVForth trial is between Ferrytoll and Edinburgh Park, a route largely or wholly on motorways. Motorways are statistically the safest roads, so this is logical. But the point of bus services is that they are the closest mass transport that provides a door-to-door service. So autonomous operation is not practical in the settings where bus travel has clear added value. Unless, of course, you see future bus operations as being about travelling between ‘transport hubs’ and out-of-town business parks.

(Tech-fans will no doubt argue ‘but what about when the tech can drive on high streets, suburban streets i.e. mix with pedestrians, cyclists, deliveries, taxis et al?’ We note that the technocrats have, for the last 10 years, said this is ‘just around the corner’, and there is still no sign it is imminent.)

The fuel efficiency of Autonomous Buses is cited as 10-15% better than those with a human driver. The reason is similar to that behind reliability improvements. Even on an electric bus, this would be a clear benefit.


This reporter is ‘not convinced’. That is not an argument for no further action. But it is a statement that of the claimed benefits of Autonomous Buses: improved safety, reliability and fuel efficiency, the first two are marginal at best. Whether fuel efficiency gains would outweigh the other disadvantages (and the cost of fitting and maintaining this advanced technology) remains to be seen.

Perhaps, therefore, any future application would be on (private) coach tour-type operations. Or on long-distance buses. It is clear there is little or no prospect of autonomous buses whizzing round city or town streets on public services in the near future.

The fundamental challenges facing the ‘local’ bus industry look to be untouched by Autonomous Buses; congestion, car-oriented transport, economic, planning and social policy. The danger is that continuing fascination with ‘new tech’ serves, as in many other cases, as a medium term diversion from addressing the fundamental issues of sustainable transport.

By all means go on a trip on the AB1 route, when the time comes. Just don’t think you’re on a transport revolution.