Jason Ford, news editor
Discussions surrounding the future of mobility often focus on vehicle autonomy, which is unsurprising given that IHS Automotive predicts global sales of nearly 21 million self-driving units in 2035.
Google is arguably one of the more high-profile proponents of vehicle autonomy, and its fleet of vehicles on public roads has clocked up over two million miles of driving in fully autonomous mode.
The company’s SUVs and prototype vehicles are reporting between 24,000 and 26,000 miles per week in autonomous mode, but Google is only one of many companies looking to make the leap into autonomy, bringing with them a raft of suppliers keen to deliver the systems that will ensure vehicle safety.
One such entity is Ford Motor Company, which plans to have a fully-autonomous ‘ride-hailing’ service on the road by 2021 and has made investments in Silicon Valley to help make this happen.
Ford’s solution will be built without steering wheels, accelerators or brake pedals, with the company contributing an undisclosed amount of funding to Civil Maps to bring high-resolution 3D mapping capabilities to the automotive giant’s autonomous cars.
In the UK, start-up FiveAI aims to bring Level 5 autonomy to vehicles through the application of artificial intelligence.
According to The Engineer, July 2016, early approaches to autonomous vehicles have required accurate 3D maps built using point cloud technology. In use, each vehicle then correlates against that map to work out where it is and establish a track to follow.
By using AI, machine learning and computer vision, FiveAI’s software will remove the need for highly detailed prior 3D mapping of environments.
So far we’ve touched on examples of vehicle autonomy occurring in the experimental domain; the proposed manner in which certain autonomous vehicles will be used; and a glimpse at the shift of focus amongst OEMs who may be looking more to Silicon Valley and regions like it for the technologies that will sell their vehicles.
Whilst we may not know what the vehicle-buying public will want from mobility in 2035, we can look to a survey conducted in Europe by Goodyear/ThinkYoung that shows 85 per cent of so-called Millennials (people aged 18-30) planning to own a car in the next years.
This demographic shows the automotive industry as having a significant role to play in a more eco-friendly future, with smart, affordable and connected cars toping the automotive wish list by 2025. Millennials are keen also on personalisation, with popular future developments that include:
- Adaptable CO2 emission levels (33.5 per cent)
- Vehicles that adapt to all terrains and weather conditions (32.8 per cent)
- Vehicles that adapt to selected fuel consumption targets (28.7 per cent)
- On-demand car services (24.8 per cent)
- Vehicles that adapt to driving style (19.5 per cent)
According to the survey of 2,564 millennials, autonomous vehicles have a role to play in smart safety, but 40.6 per cent prefer basic levels of autonomy (e.g. cruise control and anti-lock brakes). The report adds that the biggest hurdle to fully autonomous, self-driving cars in 2025 was identified as reliability (55.5 per cent), affordability (45.7 per cent) and concerns about security and privacy (38.5 per cent).
“As they have middling confidence in autonomous cars due to reliability and safety concerns, millennials want the development of autonomous cars to focus on safety and a stress-free experience,” the survey stated.
Goodyear and ThinkYoung will be taking these and other issues into their Future of Mobility 2016 conference, which takes place at London Transport Museum on October 12.
The final panel discussion will consider the pros and cons of driverless technology on our roads, and what needs to be done to make the transition as smooth for society as possible.
Autonomous mobility is an area that’s advancing incredibly rapidly, and one that we’ll be exploring in more depth in the November issue of The Engineer.
In the meantime, let us know your thoughts on the technology in the comments section below.
Two million miles is nothing.
There are 35 million vehicles in the UK. The average mileage of a car is 8,000 mile per year. I make that 280 billion miles per year.
There are 1700 road deaths per year. I make that 163 million miles between road deaths.
I ‘ve heard of two road deaths caused by self-driving cars, but what is their total combined mileage?
They are not as safe has human drivers.
I find the 85% statistic surprising and suggest it relates to a strongly skewed survey demographic.
Perhaps that should have read: “85 per cent of so-called Millennials (people aged 18-30) INTERESTED IN CARS AND planning to own a car in the next years.”
I have not heard anything about arriving at your destination of choice?, I have not encountered a SatNav that is capable of this to more than say 95% accuracy, in some specific cases near my location in N.Wales it could take you 5 miles away from where you want. Part of the problem is not only the technology it’s the vagaries of how a places address is denoted
Why the focus on young people ? At 82 my days of driving are numbered and I think hiring a car impossible. So, an autonomous car would extend my independence beyond my ability to drive myself. That is far more important than pandering to the wishes of any young person, Millennium or not. Younger people often tell me they wouldn’t like an automatic gearbox because they like to have full control themselves. I would have said the same thirty or more years ago. I conclude that it is older drivers who will lead the market for autonomous cars.
Who can afford this silly stuff?
Have we finally arrived at zero failure rates on machinery/controls? A crowded city street or high speed roadway are equally poor test cases.
Would you fly on a pilotless airplane? This has been possible for a long time but has yet to be employed.
The reason airplanes have pilots is to take over when things go wrong. Most landings are now fully automatic. Google’s 2 million miles of road testing have been done mostly on wide US roads. I know they are looking at navigation around hazards, but can it tell the difference between a salted road to an unsalted one, and an unsalted one which is wet (icy) or dry? Maybe a driverless car will just drive at 5 mph whenever the temperature drops to 3 Deg Celcius! Other non-standard examples are temporary speed limits, diversions set up at short notice due to a blocked road, flooded roads and narrow lanes with oncoming traffic. I fear that that the latter case will result in nobody going anywhere because “Computer says No!” Another example is driving into a temporary grass car park (e.g. at a showground) where you are directed by people waving their arms or just telling you where to go and park. I am sure it would do a great job of reversing my caravan onto a pitch if it knew exactly where to park it, but many caravan sites are basically just fields. Mind you, a driverless car would not recognise the location as a drivable one and so it would not take me there in the first place! I remain to be convinced that driverless cars can totally replace drivers.
“….Would you fly on a pilotless airplane?”
Reminds me of the school boy joke of the 50s
” pilotless -plane, everything so well tried and tested that opening announcement to passengers describes all the safety features and concludes: “…. and absolutely nothing can go wrong,,,go wrong..go wrong..go wrong..go wrong…..” [et tu Brute]
A human driver may have to make an instant decision whether to run over a child or pull away into increased personal danger. He has to make a brave moralistic decision in fractions of a second. Would any of todays algorithms in an automatous vehicle be able to make such decision, I would have serious doubts because there would be millions of accidents scenarios to deal with.
Excellent point, John. In this video from Stanford University, engineers are addressing many scenarios brought about by vehicle autonomy: https://www.theengineer.co.uk/video-of-the-week-the-ethics-of-automotive-autonomy/
This is an hopefully scenario, in reality, numerous studies show that 99% of humans will have an instinct reaction which is pushing on brakes and pull over, and that is what we see every day in the road. Thus, human reaction is not in “fractions of seconds”, nowhere of that. There is always at least one to two seconds, between seeing the thing and the time our brain will process and give the instructions to our members (hands and feet). Drive-less cars will have sensors which works at the speed of light and computers analysing these data thousands time by seconds, and reacts in same time. The only limit on what the car will ever be able to do is physics as they will only be able to prevent an accident or its gravity by the amount of energy the car is carrying (traducing: the speed of the car at that time, which will conditioned the time it have to stop).
Now, we will have to question what we want: a perfect system (that will never happened) or a system better than humans but that can failed also in some circumstances? The objective is to reduce the number of car accident and deaths, not to make a zero death system which is impossible by physics.