up to 1/4
of driving miles
could be covered by electric AV
A new study from the Boston Consulting Group predicts that by 2030, up to a quarter of all driving miles in the United States could be covered by self-driving electric vehicles operating in shared service fleets in cities. The experts seem certain that the car of the future will take hold over the next decade, but what will be the impact of this shift on society as a whole? Two researchers from UT Austin who have looked into the topic presented a summary of their findings at the South by Southwest (SXSW) interactive festival held in Austin, Texas in March.
Kara Kockelman, a Professor of Transportation Engineering at UT Austin, focused on assessing the consequences of the advent of tomorrow's cars in a number of areas, including road safety. Many people will share her argument that since human error is currently the cause of 90% of all accidents and 40% of fatal collisions are due to alcohol, drugs, fatigue or driver distraction, eliminating the human factor should drastically improve road safety. This will also mean pecuniary savings for the country as a whole. Professor Kockelman told the SXSW audience that the overall cost of six million accidents, entailing the death of 32,675 people in the United States in 2014, had amounted to over $500 billion.
Saving lives, saving time and saving money
Professor of Transportation
Engineering at UT Austin
I don't believe that autonomous vehicles will make traffic flow more freely. They might even make things a bit more complex.
On the basis of this hypothesis, Dr Kockelman calculated the number of lives that would be saved if a certain percentage of cars on the road were autonomous vehicles (AVs). She deduces that if just 10% of cars on the road no longer had a human being at the controls, 1100 lives would be saved, which in turn would add up to total savings of $1,390 per vehicle per year.
Moreover, these savings generate further savings. A lower number of accidents has a positive effect on traffic flows. "Accidents are responsible for close to 25% of current traffic congestion and the resulting delays. With a bit of luck, AVs will result in fewer traffic jams following collisions", reasoned the UT Austin professor. This, according to her calculations, will save time, petrol, plus at least $1,320 per self-driving car per year in other costs.
Kockelman has no doubt that AVs will have an economic impact but she nevertheless has some reservations about their overall effects, underlining: "I don't believe that these vehicles will make traffic flow more freely. They might even make things a bit more complex." After all, cars designed to be extremely cautious, which are therefore likely to travel quite slowly, might actually lead to traffic jams – at least in the initial phase. "We need to get self-driving car manufacturers to stop programming them to drive so cautiously once they’ve acquired the necessary experience, but that’ll take time", she argued.
Self-driving vehicles will travel more slowly for safety reasons – i.e. in order to minimize the impact of any accident – but also to ensure that the car enjoys a longer life. So a driverless car will probably be more expensive than a traditional vehicle. For this reason, plus also to take account of the new kinds of risks that are likely to arise, the Professor expects the cost of insuring autonomous cars to rise.
Kara Kockelman shows her conclusion to the SXSW audience
Greater mileage, more frequent trips, plus optional or facilitated parking
While the new technology promises safer travelling and, eventually, fewer traffic jams, what about parking, which today takes up a lot of drivers' time? Says Dr Kockelman: "We need to authorize AVs to park. I wouldn't recommend that cities with high traffic density have driverless private cars driving around the streets continuously. That would be a mistake." She does however agree that shared service fleets should in fact travel the streets continuously so as to free up parking spaces. Driverless vehicles could park in less expensive areas away from the city centre at a lower cost.
Kara Kockelman agrees with other experts in the field that in future cars will generally be shared. In the case study she presented, each self-driving vehicle will replace 8 to 13 private traditionally-owned cars. When people no longer have their own private vehicle, which they drive for perhaps one hour per day, the cars in use will be out on the road for eight hours a day on average, she reckons. The new mode of transport will also enable older people, those with reduced mobility or without a driving licence, to get around more easily. "Children could also be authorized to go off in self-driving cars on their own, depending on their age", the Transport Engineering professor suggested.
People won't necessarily cross the United States by road. They’ll travel less far than they would by plane but more often
Meanwhile, the switch away from petrol engines to electric drive vehicles would involve some adjustments to urban planning: the authorities will have to decide where the electric recharging hubs should be located and how many should there be. And in any case, whether shared AVs are electric or petrol-driven, they will radically change the way we move around, underlined Dr Kockelman.
Kara Kockelman foresees people in the United States making longer journeys by car than they do today, leading to a drop in the percentage of people opting to fly to their destination. She predicted: "People won’t necessarily cross the United States by road. They'll travel less far than they would by plane but more often." This will provide an opportunity to relax read, think about the future or find other things to do instead of driving the vehicle. "I've tested an AV and I have to say it's pretty boring", she told the SXSW audience with a smile.
Dr Kockelman has been working in this field for many years and has written a number of research papers on the subject. It will be interesting to see which of her predictions will be borne out and what the legislation on the subject in each US state will allow or require. This is still an area of great uncertainty, although some states are forging ahead in this field.
However, several major trends are emerging, with the state of California usually among the front-runners. The state Department of Motor Vehicles proved this once again a few weeks ago, proposing a set of new self-driving car regulations, which nevertheless have yet to be adopted. A key point is that the California DMV envisages authorizing self-driving vehicles to travel on state roads without requiring that a human being must be behind the steering wheel to take control in an emergency.
Meanwhile driverless vehicle manufacturers will have to comply with a number of conditions, including the obligation to apply to the DMV for a testing licence and to be able to control the vehicle remotely during tests. Manufacturers might well even stop building cars with steering wheels and brake pedals and pursue more futuristic designs.
In fact, as long ago as 2013 the state of Michigan authorized self-driving cars to travel on the roads without a person behind the wheel, and pilot tests have been run. However, the first state to allow testing of self-driving car tests was Nevada, in 2011, the UT Austin Research Associate reminded her audience.
At US Federal level, the National Highway Traffic Safety Administration is responsible for the regulation of autonomous vehicles. In October 2016 the agency published a set of directives on cybersecurity for driverless motor vehicles. Canada and Australia on the other hand do not have any federal legislation in this field. The European Union has not passed any specific legislation in this area either, but "EU member states are doing more or less the same thing as the US states: lots of pilot programmes and self-driving vehicle testing", Lisa Loftus-Otway pointed out. France, the Netherlands, Sweden, Finland, Germany and the UK are all currently running tests.
Meanwhile, aside from the technical requirements and safeguards, regulations are being drafted to protect citizens from hacking and invasion of privacy and to determine liability in the event of an accident. These are all fundamental issues that need to be properly resolved before self-driving vehicles take over our city streets and roads. The UT Austin researchers and the BCG report concur in predicting that this will be happening in around ten years' time.