That’s testament to the huge change seen in the automotive industry over the last two decades, from the propelling of electric vehicles (EVs) into the mainstream, to governments promising to crack down on petrol and diesel car production by mid-century, and a wealth of companies ploughing cash into the software required to produce driverless cars.
From the likes of Tesla, General Motors, and Waymo, tech companies and established automotive giants are slowly moving us towards a driverless society. But with a range of tech and legal challenges, how close are we to a world of fully autonomous driving?
What is the current situation with autonomous driving?
Car manufacturers and tech giants across the world are teaming up to deliver self-driving vehicles to the masses as quickly as possible.
Autonomous driving technology company Waymo—formerly the Google self-driving car project—has partnered with Chrysler and Jaguar Land Rover, outfitting Chrysler Pacificas and Jaguar I-Paces with Waymo’s sensors, radars, and autonomous driver technology.
So far, Waymo’s robo-taxi service has only been deployed in Phoenix and San Francisco in the US, but has logged over 20 million miles of autonomous driving without a single fatal accident.
Other robo-taxi services are taking off. In November, Cruise, the self-driving car company owned by General Motors, launched its first driverless robo-taxi in San Francisco, while the Hyundai- and Aptiv-owned company Motional is currently upgrading its Las Vegas robo-taxis to be fully driverless by 2023.
In addition to taxi services, autonomous vehicles (AVs) intended for private use are being developed. Daimler is working on features allowing automated driving in traffic jams as well as automated parking, “where the vehicle searches for parking spaces in car parks and parks itself while you already go shopping,” explains Sarah Widmann, spokesperson for the company.
Known for its Autopilot feature, which assists users with steering, braking, and accelerating, Tesla is also expanding its self-driving tech. CEO Elon Musk announced that drivers could request permission to beta test its “Full Self-Driving” technology, which includes more advanced driver assistance features with active guidance.
Not all of these new advances are on the same level of automation. The Society of Automotive Engineers (SAE) has established levels ranging from 0 (no automation) to 5 (full automation) to describe the capabilities of self-driving vehicles.
For instance, Tesla cars are Level 2. They are equipped with features that allow hands-free driving for certain periods of time and on certain roads, but drivers are always required to stay alert and intervene if necessary. Waymo’s robo-taxis, meanwhile, are Level 4. The car can fully drive itself without driver supervision under certain conditions.
Right now, according to a report by consulting firm McKinsey, the industry is focusing primarily on partially automated technology—or Level 2—with driver assistance features like automatic emergency braking or lane centering, which increases safety but still requires the driver’s full attention.
But as evidenced by Waymo, Cruise, and others’ inroads into Level 3 and above technology, a fully self-driving future is coming. So, what can we expect?
What could a self-driving future look like?
According to Mauricio Marrone, a senior lecturer at Macquarie Business School in Australia, driverless cars could lead to increased safety, better road efficiency, and more freedom for drivers.
Safety is by and large the main benefit of self-driving vehicles. 1.3 million people die in road crashes every year, according to the World Health Organization (WHO), and 94% of those fatal accidents are caused by human error, according to the US Department of Transportation.
Human drivers are flawed: they check their phones, turn towards their kids in the backseat, get behind the wheel even when tired or tipsy, and speed through red lights. Automation provides a safeguard against risk-prone behavior by detecting potential obstacles, maintaining a reasonable speed, and allowing the driver to safely take their eyes off the road.
There is currently some debate around the safety of partially automated vehicles, which may require the driver to intervene in situations the car has not been programmed to recognize. The risk here, according to Freek Vermeulen, professor of strategy and entrepreneurship at London Business School, is that drivers may become more complacent and less able to react quickly as self-driving vehicles become more commonplace.
But while technology is still fallible, the safety benefits greatly outweigh the potential safety risks that come with defective software or delayed reaction times, Freek argues. “The technology of driverless cars will improve, and they will become safer and safer, and people losing the skill to drive will become a smaller and smaller problem,” he explains.
Not only would roads populated by self-driving cars be safer, but also less congested. “If all cars on the roads were autonomous vehicles, speed limits could be removed, and cars could travel at faster speeds safely,” thereby increasing road efficiency, Mauricio argues.
If they no longer have to focus on the road, people would be able to occupy their time reading, working, watching TV, or talking with fellow passengers—creating a new sense of freedom that would change the way we commute and travel.
But while AV manufacturers and their investors envision and promote a shiny future of safe roads, reduced traffic, and happy (non)drivers, Mauricio believes we should maintain some skepticism. The cons of self-driving cars include the serious risk of hacking as cars become reliant on software and the loss of jobs in the transportation services industry, like taxi, bus, and truck drivers.
Mauricio also casts doubt on the claim that driverless cars would automatically reduce congestion. “Imagine doing a brief visit to a friend. You instruct your AV to drive around the block until you are ready to be picked up. All of a sudden, your ‘quick’ visit to your friends turns into an all-night event,” Mauricio explains.
Now, imagine that everyone else does the same thing. Cars that would otherwise have been parked are cruising around the block, for hours on end. While this may not translate directly into negative environmental impacts if the cars are electric, it would most certainly worsen road congestion.
Ultimately, what the future of self-driving cars looks like will depend on how manufacturers, users, and policy-makers decide to shape this emerging technology.
How do we make self-driving cars the norm?
The widespread adoption of driverless cars hinges on three factors: customer demand, technological development, and the regulatory landscape.
Self-driving cars need to gain legitimacy. Manufacturers are faced with the task of earning people’s trust in driverless cars—which will then translate into demand.
One way of acquiring legitimacy is to target one demographic first; in this case, the wealthy. Freek of London Business School argues that driverless cars will first be marketed to a relatively small, exclusive market, and progressively become more and more widespread as production scales up and prices drop.
In 2020, Mauricio of Macquarie Business School co-wrote a research paper looking at people’s perceptions of autonomous vehicles to figure out what could contribute to increased uptake.
Mauricio and his colleagues found that what mattered most to people were reliability and safety, price, and ease of use. Showing how self-driving cars could improve mobility for often neglected communities—for instance people with physical disabilities who cannot drive and have limited public transport options—could also help create demand.
Manufacturers have to take these factors into account when designing and subsequently promoting their vehicles.
The technology of driverless cars is fast-evolving. Established companies and startups across the world are using engineering and computing expertise to develop more sensitive sensors, more accurate radars, and more sophisticated software.
“From a technical point of view, the most difficult part in developing automated driving systems is finding the perfect tuning between customer benefit and safety,” says Sarah from Daimler.
AVs from various companies are being thoroughly tested to ensure that all scenarios are accounted for, from navigating around unexpected obstacles on the road and making way for emergency vehicles, to stopping at an intersection if a traffic light fails.
But developing algorithms to ensure passengers’ safety is not simply a fact-driven process. It also involves complicated ethical questions about who, in the case of an accident, deserves to be saved.
For example, five years ago, Mercedes caused outrage when they announced that its AVs would be programmed to protect the lives of their occupants, even if that meant sacrificing pedestrians.
Edmond Awad, a lecturer in the Department of Economics and the Institute for Data Science and Artificial Intelligence at the University of Exeter, led the development in 2016 of Moral Machine, a website that gathers human decisions on moral dilemmas created by driverless cars.
The website presents a simple scenario akin to the famous ‘trolley problem’. A group jumps in front of a driving car; you can either hit them and kill them, or swerve to the other side where there is either a barrier or another group of people, and if the car hits the barrier, it will kill its passenger.
Multiple variables are included in the experiment, both structural (the number of pedestrians, whether they are crossing legally or illegally, for example) and potential victims’ attributes (human or pets, age, gender, social status). Overall, there were 20 different characters presented in randomly generated scenarios.
In the first 18 months of the launch, the website garnered over three million users who participated in over 40 million dilemmas, producing enormous amounts of data that Edmond and his team analyzed.
They found that overall, people preferred to spare females over males, young over elderly people, humans over pets, and pedestrians over passengers.
Interestingly, those preferences varied between regions. For instance, in Central and South Asian countries, the preference for young over old people wasn’t as pronounced as in Western societies. In Germany and Japan, where respect for the rules is strong, participants in the experiment found it more acceptable to spare the lawful pedestrian over a jaywalker. Depending on how individualistic a society is, and how strictly it abides by the rule of law, who gets spared varies quite widely.
The fact that different cultures have different ethical perspectives has important implications for manufacturers of driverless cars. Most driverless technology is being developed in the West, which means algorithms will most likely reflect Western rather than universal ethics— like preferring to spare the young over the old, and the lawful over the unlawful.
But right now, Edmond argues, manufacturers are not being transparent about the values that they are embedding in their machines. “Companies are not really incentivized to really care about these issues,” he says. “They see it as a waste of time or resources, so any action has to come through policy, through some kind of regulation, otherwise we can’t really count on companies to do that on their own.”
Regulation more generally, not just over driverless cars’ ethical dimensions, will determine if and when we get to a self-driving future.
Who is liable if the car kills a pedestrian? The manufacturer or the driver? Who gets to have a license? Those are important questions that regulators and lawmakers will have to answer.
But the law is notoriously slow. “When cars came about, which eventually replaced horse-drawn carriages, the whole legal system had to adapt,” Freek from London Business School explains. “The law has to gradually figure out that this is something new. But that will take a long time.”
The UN Economic Commission for Europe and several governments— like Germany’s, which opened up its road traffic legislation to SAE Level 3 systems in 2017 and Level 4 systems in 2021—are actively building up a regulatory environment for driverless cars. But requirements remain unclear, especially when it comes to vehicles with higher degrees of automation.
Getting the law to catch up with the technology is arguably the biggest hurdle. Regulations around Level 2 features are more easily defined, but regulators may struggle to establish requirements for Level 3 features and higher.
That’s why progress will be incremental, with partially automated driving becoming more and more prevalent, and technology progressively getting better, until we get to a truly self-driving future.
In the meantime, important questions about what customers really want from their self-driving cars, and how they would like them to behave in critical situations, remain to be answered.
BB Insights draws on the expertise of world leading business school professors to cover the most important business topics of today.