Driverless transport is coming. In fact, it is already here. There are already travel solutions where a human driver is present essentially for backup purposes. Then there are examples where the driver is completely absent from the vehicle. In both cases, it is the advanced computer systems, with elements of AI, which take control.
In 2016, the National Highway Traffic Safety Administration (NHTSA) in the U.S. adopted the Society for Automotive Engineers' (SAE) revised six-level scale for 'driver assistance technology'.
Where 0 represents today's standard of fully human controlled vehicles, 5 stands for entirely automated driving with no human interaction required, and essentially not even the need for a steering wheel. This scale enables an appropriate autonomous-automotive classification, as well as the integration of driverless cars into vehicular legal systems.
Rush to be first
Over a billion motor vehicles are on the road today, so it is little surprise that car and tech companies alike are rushing to be the first to launch automated cars into the mainstream market.
Google announced that they were developing driverless technology on the official Google blog, as far back as 2009. However, things did not really pick up speed until February 2015, when competitor Uber announced plans of their own to develop driverless automobiles.
Google Ventures was actually an early investor in Uber, investing $258 million back in August 2013. As of 2015, Google and Uber were racing the development of their driverless cars to the finish line.
Google's self-driving car project became Waymo in December 2016, and they continued to develop and test driverless technologies.
Waymo vs. Uber
On 5 February 2018, Waymo and Uber appeared in court, involved in a lawsuit where Waymo accused Uber of stealing thousands of documents containing trade secrets around the design of driverless technology. The two tech heavyweights reached a settlement four days into the trial, on 9 February 2018.
Uber agreed to give 0.3% equity to Alphabet Inc, Google's holding company. At the time, Uber was valued at $72 billion, which results in about $245 million in equity for Alphabet Inc. There is also an agreement from Uber to work openly with Alphabet, in order to ensure that none of Google's confidential technology makes its way into an Uber design.
The danger of driverless cars
On 18 March 2018, one of Uber's driverless cars collided with a pedestrian, 49-year-old Elaine Herzberg from Tempe, Arizona. Herzberg later died of her injuries. This was the first time a pedestrian had been killed in an incident involving a classified autonomous vehicle. Her death has been seen as a strong argument against driverless automobiles.
The incident, although tragic in itself, is one of 40 000 deaths caused by cars each year in the U.S., and one of 1.25 million deaths worldwide. The world is no stranger to the dangers of the roads, but is it worse to hand over our lives to an artificial intelligence, instead of accidents occurring as a result of human error?
Man vs. machine
When Walter Huang died on 23 March 2018, as a result of his Tesla Model X car hitting a highway barrier, there was much commentary and criticism on all sides. Opinions were divided over who was truly at fault.
Tesla stated that the level 2 car, while on autopilot, still required the driver to have their hands on the wheel. They claim that Huang did not have his hands on the steering wheel for six seconds, despite several warnings from the car that he should do so. However, this argument might be considered somewhat shaky coming from a company whose aim has been to convince the public that an Artificially Intelligent driver is safer than a human one.
Public testing ceased
Uber and Nvidia both immediately ceased public testing of their car programmes following the incident, potentially destabilising their respective deadline calenders. However, there appear to be few widespread effects. Perhaps due to the hype surrounding the tech, as well as the contrasting statistics regarding risk and safety of autonomous road vehicles, the industry is still set to continue towards the ambitious plans to remove humans from the driver's seat.
The majority of cars being tested in public are level 2-3 autonomous vehicles on the SAE scale. This means the vehicle is conditionally automated, and a human presence is necessary to take control when needed.
Regulation is complicated
Regulators have to work with the emerging technology itself to understand how to regulate it. This means baby steps, guided by careful observation, public/political opinion and input from business lobbyists. This applies the world over, and differing legislative bodies are slowly beginning to respond to the oncoming tide of driverless vehicles that, if the companies are to be believed, is only years away.
This map by Bloomberg Philanthropies shows every city on the planet where driverless cars are being driven, or where they are currently drafting plans to do so in the coming months and years. There are 66 cities running trials at this time, five of which are in Scandinavia.
Gothenburg is hosting Volvo's flagship Drive Me project. They have opened up a circular route in and around the city where they hope 100 participants will be in the driver's seat, while not actually driving the Volvo hybrid XC90 SUV.
This project was set to release level 4 vehicles in Sweden, UK and China over the next two years and started in December 2017, with two families based in Gothenburg already using level 2 vehicles.
That same month, however, Volvo pushed back the schedule and scale of the project. Only level 2 vehicles are to be used in the initial roll-out, and the first phase is set to finish in 2021.
A well-timed decision
Volvo's XC90 model is the same as the car involved in the fatal incident in Arizona. Volvo also uses the same Nvidia self-driving chips in their own autonomous vehicles. With those things considered, the decision to slow down and scale back can be said to be timely.
The autonomous automotive industry appears relatively unscathed despite this, and numerous other, setbacks. Scaling back the ambitions of all companies involved seems like a sensible option if the aim is to prevent negative press from crushing the industry before it has really started. It can also be considered wise from the perspective of human safety.
Aiming too high?
Jack Stewart of Wired.com makes the case that the autonomous car industry needs to “aim lower” if it is to succeed in winning over both the public and lawmakers.
Stewart makes a good point. Whilst making money is at the foundation of most companies, and more money more quickly ensures rapid technological and marketing advances, if this besmirches the reputation of the technology itself, then the effect is too damaging.
In order to win people over, the focus needs to remain on increasing safety, not just for passengers but for all civilians.
Improve public opinion
There are other ways to potentially improve public opinion. Stewart suggests that companies could offer a free taxi service in the drunken hours of the night, in order to help vulnerable people reach home safely. This sort of initiative might just secure the sought-after good press that the industry needs. Such a project could even generate loyalty from young adults, who might be their main target audience.
The argument remains as to whether vulnerable people really are safer in a vehicle controlled by a robot.
If private companies never do manage to convince the public, and the authorities, that driverless automobiles are safe, then it is unlikely that a government or authority will take them on either. Safety must come first and a few more deaths in the testing phase could be disastrous for the future of public and private initiatives alike.
Despite all this, many do believe that driverless cars are coming. Level 2 automation is the natural testing phase, but transport with level 4 automation has been around for many years within closed guideline networks. Trains guided by computers have been shuttling passengers to and from Heathrow's Terminal 5 since 2011, running with level 4 autonomy. Multiple similar projects are running across the globe with many more on the way.
Scaling these systems up and integrating them into public transport networks is the next big step. The likelihood is a continuation of public transport automation. We will certainly see the introduction of more autonomous trams, trains and buses before cars.
This is an area national governments are still heavily investing in the world over. They are attracted by the reduction in human error, as well as the increased efficiency of computer-controlled public transport. There is also the environmental benefits, as most public transport options involve hybrid or electric vehicles.
The next step should be a softly, softly approach. More trials would give companies time to strengthen the existing tech, while simultaneously building public support. This should lead to reassuring regulators and attracting political and economic interest.
Most likely we will see taxi, shuttle and delivery services working within pre-defined and regulated boundaries over the next few decades. However, it would seem that buying a car and expecting to be chauffeured to wherever you choose, is not on the table just yet.
The costs are still too high. Legislation regarding safety, legal responsibility and transport infrastructure are still a long way off. It is believed by many to be possible, though, and with estimates of $7 trillion global gains, it is less a case of if but when.
We believe that information should be free and will therefore never put up a paywall.
If you like reading our reports about the Scandinavian business scene and would like to donate towards the upkeep of the site, we would be very grateful. Click here to donate.