SovTech is now Scrums.com!!
Read more here
Autonomous Vehicles: So much more than self-driving cars
The idea of self-driving cars has ceased to be a futuristic notion; they are a reality, with mounting excitement, concern and investment potential as the capabilities of driverless vehicles grow. Speculations abound of what the future may hold for our roads and road-user cultures ranging from something close to a utopia – where no one owns their own cars, transport is at once private and public, and is cleaner, safer and more efficient – to a panic-mongering rhetoric in which autonomous vehicles (AVs) are forced to make moral decisions, millions lose their jobs in transport and related industries, urban sprawl becomes even more widespread and cyber terrorists have the ability to cause mayhem on cities’ roads.
International automotive standardisation body, SAE, has classified the range of automated driving, from Level 0 where an automated system issues warnings to the driver but has no control over the vehicle, to Level 5 where no human intervention is required. This is the realm of the self-driving car.
AVs employ various systems of artificial intelligence to anticipate, recognise and navigate threats, objects and the road. In order to obtain as accurate a positioning as possible, a combination of GPS signals, tachometers, altimeters and gyroscopes is used. LiDar (light detection and ranging), ultrasonic and radar sensors fitted to the roof, near the wheels and on the front and rear of the vehicle bounce signals off the surroundings. And video cameras detect road signals – signs, traffic lights, pedestrians, and other vehicles. All of this information is interpreted and acted upon by a central computer which enjoys exclusive control over steering, acceleration and braking. The software must understand all formal and informal rules of the road, and be able to rapidly analyse a large amount of data.
Driverless vehicles have been in operation for some time now. Google, as Waymo, has been experimenting for 8 years and their AVs have accumulated close to 5 million kilometres on public roads. AVs have been operational without incident at certain terminals at Heathrow airport for 5 years, and Tokyo plans to use AVs at the 2020 Olympics. The question is not whether or not AVs are possible but the complexity of road-user situations which they are capable of navigating safely and efficiently.
With all their accumulated real-world driving experience, Waymo’s cars are able to negotiate some of the more unpredictable aspects of the road including dodging jaywalkers, braking for animals, pulling over for emergency vehicles and interpreting signals from construction officials.
The hope for this technology is that it significantly reduces road accidents and the associated injuries, fatalities, and costs, that it drastically eases congestion, increases the mobility of marginalised groups such as disabled people and the elderly, eliminates the need for parking space in cities and enables the development of a vast sharing economy among a host of other predicted or prospective benefits.
Of course, as with anything harbouring revolutionary potential, there are ample uncertainties (besides technological challenges) such as the myriad ethical questions around potential failures in safety, the changes that AVs would likely necessitate in infrastructure and legislation. These concerns are adjacent to the driverless cars themselves, giving rise to the question: if AVs are to be introduced at large scale, what do we need besides the cars?
We may have the technology for AVs, but in order to make them viable, there is a range of other infrastructures that need to be in place.
For a start, in order for cars to be able to navigate themselves, every single road needs to be mapped in countries where AVs become available, and these maps need to remain up to date. Digital cartographers like Google Maps, Garmin and TomTom will need to up their game to fulfil the needs that market-ready driverless vehicles will generate.
According to the World Economic Forum, it has been estimated that each AV will produce and use around four terabytes of data per eight hours of driving. And the cars’ computers need to be able to process such vast amounts of data and act upon it without delay.
These prerequisites demonstrate that autonomous vehicles, if they are to reshape our roads, necessitate a transformation in the underlying digital technology and the management of data. Driverless vehicles may signal the need for clouds everywhere – distributed platforms that are able to coordinate between data centres and maximise on computing capacity.
A World Economic Forum reporter muses: “Imagine a world where there are clouds everywhere; where the devices we use are not just networked but share the burden of huge volumes of data and instantaneous decisions, where software is seamlessly managed across infrastructure wherever in the world it resides”. These are the kinds of foundational advances in information technology that will provide the preconditions for self-driving cars and smarter cities.
Besides this transformation in the digital infrastructure, AVs necessitate changes in the built environment to maximise the effectiveness of roads on which cars drive themselves. Intersections need not have traffic lights or road signs, needing instead programmes that communicate directly with the vehicles. As driverless cars become increasingly connected, so cars with drivers may become more displaced on roads, especially within the bounds of busy city centres or high-speed roads.
Legislative infrastructure is another area to which AVs will revolutionise, and this one may prove to be something of a minefield. There are two particularly controversial areas which will require thorough legislative foundations. The first is privacy – an issue which may arise from the aforementioned connected-ness of AVs, where they may become another data-gathering source on individuals’ movements and preferences as well as devices that record voice and video. The second area is that of liability. However brilliant the technology, accidents will happen. This is already a question that innovators face after the fatal accident of 2016 in which a Tesla Model S. in Autopilot ploughed into a truck. Legal frameworks will need to be sophisticated and yet flexible in order to negotiate through the maze of ethical questions cases such as this raise.
Perhaps more important than any of the above-mentioned transformations fundamental to the implementation of AVs on our roads is the transformation in the way the public thinks about driverless cars, so that people may be convinced of the great potential benefits, especially in the areas of safety and efficiency that driverless cars offer.
The question does beg to be asked though: is the hype around this exciting technology a case of a game-changing and under-estimated revolution, like the advent of the internet, or is it something more akin to the predicted death of the book, a particularly hardy remnant of the analogue world?
Most likely, if autonomous vehicles can contribute to healthy streets and city spaces, and productive and happy people, then they may just be all that they are cut out to be!
