Robocabs: How Long Before you Ditch your Car for a Driverless Electric Taxi?

Guest post by Richard Brooks.

Trend-spotters may have declared the car is dead for 20-somethings in central London or Paris but among the rest of humanity sales of the ubiquitous gas-guzzler continue to climb. It seems however environmental we may wish to be, owning a set of wheels is just too convenient to give up.

But maybe not for long. A more radical solution to tackling climate change is proposed in a paper in the journal Nature Climate Change. As the industry inches towards self-driving cars isn’t it time to consider trading in the family wagon and simply hail a lift?

The argument is seductively simple. We can’t do without cars. No transport alternative offers the same flexibility, personalised comfort and sense of control. But automobiles are spectacularly inefficient in terms of environmental and personal cost. Cars depreciate rapidly, are left idle most of their life-time, eat through fuel while seats are often left empty and they demand no end of real-estate to be parked on.

If any field is overdue a wave of disruption it’s personal car ownership. Car-sharing schemes such as BMW’s DriveNow and on-demand services such as Uber give us a glimpse of the benefits cloud computing can bring to fleet management. Autonomous vehicles (AVs) however, could transform this field.

Ditching the driver

Autonomous taxis would be more expensive than a normal car, but with no driver the running costs would be substantially lower. Operators would therefore be likely to run large fleets and would keep their cars on the road for as long as possible – meaning that in areas of high population a vehicle would never be too far away.

These vehicles would be different. Just as electric cars have increased interior space by doing away with the engine, dispensing with the steering wheel, gear stick and other controls would free up room for extra people and their belongings.

This also links to the potential for what the paper’s authors, two Berkeley scientists, call “right-sizing”. If a wide range of vehicles are in constant circulation, people will be able to order whatever suits them: a people carrier to take the kids to school, then a sporty coupé for the commute to work. This would mean more choice for the individual, but with a knock-on benefit for the environment through reducing the number of large fuel-hungry vehicles with only one occupant.

It points towards another expected benefit of autonomous taxis. If fleet operators want cheap and easy re-fuelling (to lower those overheads) then fully-electric vehicles make sense. Whereas single car-owning individuals are likely to be concerned about vehicle range, taxi firms smooth out journey and fuelling requirements over an entire fleet, enabling them always to take advantage of the cheapest electricity.

Autonomous vehicles could revolutionise driving. Sensors linked to sophisticated electronic control units would enable cars to anticipate and to respond to risk faster than a human driver. Breaking distances could be reduced, enabling convoys of cars to reduce aerodynamic drag and use their power more efficiently.

Heading to the mainstream

A great vision for the future, one might observe, but isn’t this all a bit sci-fi? Well at this stage, the answer is probably “yes”. Vehicle automation has of course been with us for years and is actively used in high-risk fields such as bomb disposal, mining or submersibles, not to mention the release into the world of robot vacuum cleaners. But these are relatively controlled environments to operate within, and a world away from the complexity of fast-paced, complex city streets.

As it stands there are substantial limitations on autonomous vehicles. Cars are reliant on processing, in advance, extensively mapped environments – making any spontaneous deviations from a route difficult. Snow, heavy rain or bright light can cause havoc with sensors, and “reading” visual symbols such as a police officer signalling would be beyond most current processors. Add to this, the cyber-security risks inherent in all connected vehicles and it becomes plain just how much work developers have to do.

Even if these challenges can be overcome in any realistic time-frame the usual drawback of car-sharing services would still apply to autonomous vehicles. Being able to hail a taxi when you want it, never mind in the form one desires it, requires a huge fleet to be operating in densely populated area. We’d also need advanced GPS mapping and tracking systems and large out-of-town charging bays – because although more autonomous taxis would mean less parking needed overall, the fleet as a whole would likely seek cheaper overnight charging at a similar time.

Yet what was once science fiction has a tendency to move into the mainstream. Carmakers such as BMW and Tesla have been acquiring driverless technology, as have tech giants Google and Apple – not to mention the car-and-driver-for-hire firm Uber. Indeed one doesn’t need to be an early adopter to see driver assistance and connected technologies becoming commonplace. Crossing the hurdle to fully autonomous vehicles will be a major challenge, but once it has been credibly achieved (and credibility is everything), a whole new market may be awaiting the pioneers.

Ultimately this is the point. From San Francisco to Moscow, from Seoul to mighty Coventry (where the UK’s largest driverless car test is taking place), there is a digital infrastructure being laid down and growing expectations among the younger generation of ever more connectivity. Autonomous taxis may not be appropriate everywhere or for everyone, but they do have the potential to capture a large slice of interest in urban centres and, because the capital costs are borne by the transport operators not individuals, there is lots of potential for the business models to spread quickly.

By 2030 our cities, not to mention our cars, may start looking very different. It might be worth thinking about grassing over your driveway.

Originally written for ‘The Conversation’.



Ryan Walker