The road ahead
Reading Time: approx. 10 min. The road ahead – Where is autonomous driving taking us?Plotting a route to 2020 and 2030
Let’s go! The topic of autonomous driving is currently having an equally electrifying effect on the automotive sector as it is urban planners, car drivers and mobility service providers. In California alone, 50 companies now have a license to conduct autonomous driving tests and the state has allowed self-driving cars with no steering wheel or pedals on its road since April 2018. In the same year, city administrators in Beijing gave the green light for tests with self-driving cars to be conducted.
However, there have also been frequent reports of software errors, fatal accidents, delays in development and uncertainty over what exactly will be possible, and when.
In other words: It’s probably time to take a look at the road ahead, from both a short-term and longer-term perspective: How will we be getting around in two years’ time? And where is the journey taking us as we look toward the year 2030?
It’s a journey with two stages and two pit stops.
STAGE 1: Destination 2020The self-driving car will affect virtually all of humanity. Which is why it will be a huge business.Larry Page, CEO, Alphabet Inc
So where are we right now? General Motors hopes to put self-driving taxis on the road in the USA by 2019. Volkswagen is planning to install mobility fleets with self-driving vehicles two years later in several US cities – and perhaps also in Europe. In Germany, the box-shaped electric city bus known as the e.GO Mover is set to be delivered from mid-2019.
It seems as though automated mobility is just around the next corner. In reality, it is still some distance away – and the road there could be bumpy.
“Ninety-eight per cent of driving is just following the dotted line. It’s the other two per cent that matters.” Burkhard Bilger, “Auto Correct”, The New Yorker.
Here are three of the many challenges that autonomous vehicles still have to overcome:
01 Autonomous vehicles detect the outside world through a combination of radar, laser and camera. Algorithms form a picture of the surroundings based on the data collected. But these surroundings are enormously complex. For example, it is still difficult for the systems to identify that a plastic bag blowing across the road is a harmless object. And the opposite can happen too: In May 2016, a Tesla Model S was involved in a fatal accident when the system mistook a semitrailer for an advertising sign.
02 What happens next in the chaos and confusion that we know as our roads? What about factors like snow, rain, darkness, unsecured accident sites and cars turning the wrong way down a street? Our human brains can deal with exceptional circumstances like these, but robotic drivers still lack intelligent understanding. “It’s no use having a fair-weather system that works in Phoenix or Houston, but not in Chicago, New York or Hamburg,” Johann Jungwirth, Executive Vice President of Mobility Services, Volkswagen. The major challenge for computer control of cars is to get to grips with these exceptional events.
03 Another problem area is differing regulatory requirements. In the USA, companies like the Alphabet subsidiary Waymo, as well as Tesla, Lyft and Uber, are able to consolidate their leading position while being largely unhindered by national regulations. In Europe, it’s far more complicated. “The straightforward approval procedures in the USA make it possible to have test fleets of several hundred vehicles – many times what is possible in Europe,” says Wolfgang Bernhart, Partner at Roland Berger.
FIRST PIT STOPWhy are we on this journey at all?
In light of the current obstacles and difficulties, the enormous benefits promised by automated driving can sometimes be overlooked. Some of the main plus points are:
Safety Around 1.2 million people around the world die each year in traffic accidents. About 90% of these fatalities are caused by human error. An automated chauffeur that avoids these errors would have the potential to go down as the biggest life saver in recent history. For example: According to a study by the University of Michigan, the human eye can see a maximum of 76 meters in the dark. By contrast, the radar systems used in autonomous vehicles (and primarily LiDAR systems in the European and Japanese markets) detect objects up to 250 meters away, even in the dark.
Living standards In many major cities, some 30 to 40 per cent of inner-city space is reserved for parking. If cars parked somewhere on the outskirts instead of outside your door, or if they were constantly on the move and could simply be ordered when needed, up to a third of the urban area would suddenly be freed up and could be used for parks, playgrounds, homes and offices.
Mobility for all Elderly and disabled people would no longer be excluded from the personal mobility offered by a car. And if you wanted to go out for a drink in the evening, you could; after all, you would have long since handed in your driver’s license anyway. Even children could be safely chauffeured to their destination.
Environment As driverless vehicles travel more efficiently through traffic, air pollution and the frequency of traffic jams would both be reduced – unless automated driving became so attractive that even more people switched from bikes and buses to cars.
Quality of life The average person spends 37,668 hours in the car over the course of their life – time that could be much better spent on other things, like working, listening to music, watching a movie or sleeping. And all while on the move, of course.
New business models One example is passenger transport. General Motors calculates that a conventional taxi earns an income of around 30,000 US dollars over its service life. A Robo-Taxi that can operate around the clock with much lower personnel costs would quickly generate a revenue of hundreds of thousands of dollars per vehicle.
Autonomous driving will turn mobility on its head just as the automobile did a century ago.Mario Brumm, Ibeo Automotive Systems
STAGE 2: Destination 2030If we look at the medium term, we can identify some interesting options.
• The time we spend in the car will become a commodity that can be traded and negotiated. Some media companies are already in negotiations with car manufacturers over the price at which their media could be offered in vehicles. Employers could include commuting time as part of working time; after all, it would be possible to work while traveling.
• Everyone can drive. If nobody needs to pass a driving test before driving: Who is allowed to drive automated vehicles? Children? Drunks? Anyone?
• Nobody needs to own a car to be able to drive one. This raises the question of who cars will actually belong to in the future: themselves?
• Cities could be built at a higher density level – in a similar way to Singapore, where car sharing is strongly promoted. In turn, the neglected outskirts could gain a new purpose as a gigantic parking lot for autonomous vehicles.
• Speaking of urban planning: As human-driven vehicles one day disappear from our cities, so too will traffic signs – after all, the robots don’t need them.
• It’s worth mentioning that autonomous driving also opens up new, less positive possibilities for criminals (self-driving getaway cars) and terrorists (car bombs that search for their target like. drones)
We always overestimate the change that will occur in the next two years and underestimate the change that will occur in the next ten.Bill Gates
SECOND PIT STOPWhat exactly are we talking about?
Not all forms of automated driving are the same. European and American authorities classify the concept in six levels:
• Level 0: “Driver only.” The driver controls the vehicle.
• Level 1: Driver assistance. The driver is supported by assistance systems such as adaptive cruise control.
• Level 2: Partial automation. Assistance systems take charge of parking, staying in the lane, acceleration etc.
• Level 3: Conditional automation. The driver does not need to continuously monitor the system but must be able to take control within a warning period.
• Level 4: High automation. The system takes permanent control of the vehicle. The driver is only there to take over if the system fails.
• Level 5: Full automation. No interventions are required, except to specify the destination and start the system.
But even if we look further ahead, some important questions remain. What will become of the automotive industry? What about the joy of driving? What impact will changes to vehicle mobility have? And when exactly can we expect the revolution in automated driving to happen?
We now have to measure ourselves against the Googles and Apples of this world, and the Teslas, and be better than them.Dieter Zetsche, Chairman of the Board of Management of Daimler AG
What will become of the European automotive industry? A better question is: What will become of the automotive industry as we know it?
“Can you imagine a customer who specifically orders a Mazda, Opel, Peugeot or Seat?” asks consultant Thomas Ridge (Ad Punctum). “So what would then be the need for 25 different brands?” Ridge therefore predicts a “mass extinction” of vehicle manufacturers.
Silicon Valley analyst Ben Thompson is similarly pessimistic. In the age of autonomous driving, he says that private ownership of vehicles will become a “massive waste of resources” for end users. For the manufacturers, this means that they will become “commodity providers selling to bulk purchasers, not dissimilar to the companies building servers for today’s cloud giants.”
There may also be some huge geographic shifts. Today, around 15 per cent of cars are still produced in Western Europe. By 2040, the major decision makers in the automotive industry expect this proportion to fall below 5 per cent. It is hard to imagine that this massive shift would occur without the loss of any jobs.
What’s more, the newcomers to the industry are each sitting on substantial war chests. Google, for example, has been working on driverless cars since 2009 and is worth as much on the stock market as the ten largest automobile manufacturers put together. Studies have shown that by 2030, the Alphabet subsidiary Waymo could supply around 60 per cent of the operating systems in autonomous vehicles. This would make this digital provider the central customer interface, giving it a lucrative share of the value-added chain in the field of mobility, explains Marcus Wieland of the Intelligent Mobility department at MHP.
If you think long term, there could be just a handful of providers of self-driving systems left over: One or two in the US, one to two in China, and one or two in Europe. Johann Jungwirth, Executive Vice President of Mobility Services at Volkswagen
What else will change, economically and otherwise?
There are a host of new business cases across the automotive industry.
One recent example comes from 3M: The US company, generally more well-known for its adhesive tape and post-it notes to date, has developed a technology that can apply machine-readable information to traffic signs. The codes can be read by infrared cameras in vehicles. Through a cloud database, these codes can be used to store the latest information, such as “Accident around the next bend” or “Ice in 700 meters.” The system should be ready for the market in about three years and could help to mitigate a key weakness of autonomous vehicles – the occurrence of unpredictable events.
“It’s really easy to do things with test vehicles that have luggage compartments stuffed with technology. But to do that it in series-production vehicles, and at a reasonable price, is something else entirely.” Mario Brumm, Ibeo Automotive Systems
When can we expect the autonomous revolution to happen?
Not for quite some time – if it ever happens at all. The transition from human to autonomous driving is likely to happen in evolutionary stages. During this process, more and more tasks will be taken away from drivers by assistance systems.
However, commercial vehicle fleets are likely to be automated before any significant numbers of private autonomous cars are on the road. These vehicle fleets will include garbage trucks, which will no longer need drivers and can travel at walking pace. Or pickup trucks that drive to their destination at night when the roads are clear, like robotic mules, and are standing ready outside the premises when work starts in the morning.
Another stage in the transition may involve geofencing. As city traffic is incomparably more complex to handle than traffic on long-haul routes, it may only be possible to approve autonomous systems for the latter. In this case, drivers would navigate the city traffic themselves and then switch to the autonomous system once they reach the outskirts.
“Without self-driving cars, we won’t have a business anymore.” Travis Kalanick, former CEO of Uber
Will we ever be able to get used to autonomous driving?
Yes. And this will probably happen surprisingly quickly, just as it did with robotic lawnmowers and cell phones (both were just toys for geeks and posers to begin with before they suddenly became part of our everyday lives), which have become ubiquitous. And it also means that we would:
• Send an autonomous vehicle to pick up pizza
• Trust it with our children let it drive them to a football game
• Share a self-driving taxi with people we don’t know, without any problems
• No longer get in on the driver side automatically, even if we are traveling alone
You can think of the car as a server on wheels.Brian Krzanich, CEO of Intel (up to 2018)
But what about the joy of driving?
Good question. The joy of driving is likely to shift over time: from the fun of driving a car to the pleasure of relaxing and doing all manner of other things while on the road. And vehicles with an internal combustion engine will probably still be on the road for decades – so if you have petrol in your blood, you will still be able to put your foot down and get it pumping.
But what about after that? “The joy of driving will become a private pleasure,” believes Mario Brumm from Ibeo Automotive Systems. “Just like people who used to enjoy horse riding and set up a racing stable once cars took over from carriages, one day you will only be able to drive a car yourself on private race tracks.”
The joy of driving will also involve a reduction in the collateral damage caused by individual mobility. “Autonomous electric mobility will make transport in urban centers cleaner, safer and more efficient,” believes Prof. Günther Schuh, Chair of Production Engineering at RWTH Aachen University and co-founder of electric vehicle manufacturer Streetscooter. “But it will make us drive more, not less. Autonomous driving means there will be no conventional stops, timetables or shutdown times anymore. Instead, we will have transport on demand. In the center of the city, individual transport will be largely abandoned. People will be shuttled around by autonomous buses.”
“The race in the networked world is completely open.” Volkmar Denner, CEO of Robert Bosch GmbH
How quickly are we progressing?
The answer is that nobody knows. Chief Digital Officer at Volkswagen, Johann Jungwirth, believes that machine learning, artificial intelligence, cloud computing and simulation capabilities will continue to grow exponentially and accelerate development.
But Prof. Schuh says that “we still need 12 to 15 years to really be sure about how to master autonomous driving. We will only be able to enter this new world of autonomous driving for real if we can be completely certain that autonomous vehicles can drive us around much more safely than a very good human driver would ever be able to. The training period – or in other words, the learning phases we need to go through before we can really master autonomous driving – will take several years.”
One thing is for sure: Around the world, development will proceed at different rates depending on the legal situation, the openness of the markets and the incentives that countries offer for autonomous driving.
Conclusion: A traffic forecast
What will happen after 2030? According to a study by consulting firm Oliver Wyman, partially and fully automated vehicles could make up between 20 and 35 per cent of global vehicle production by 2035. But will that really be the case?
Here are six theories about the distant future of automated driving:
01 Even by the fourth decade of this century, there will still be far more human drivers than robotic vehicles on the roads.
02 Autonomous cars will not significantly change our world until they have surpassed the critical mark of at least 20 per cent. This is because urban planning, traffic regulations and everyday transport will still mainly be based on human drivers.
03 The living room on wheels often seen in simulations of the future, in which passengers can lounge comfortably in an armchair or even lie down while their robotic chauffeur takes them to their destination, will be a long time coming. After all, as long as human drivers are also traveling on the roads – which will be for a very long time yet – the vehicle occupants will need to be prepared for a crash scenario. And that means sitting in a predefined position with a safety belt and airbags.
04 Driverless mobility will become cheaper. Autonomous shuttles will be available at a price of 10 to 12 cents per passenger-mile. By contrast, a driver-operated MINI costs about 55 cents per passenger-mile.
05 Increasingly intelligent assistance systems will gradually remove the need for us to drive cars ourselves. This change will give us time to get used to driverless mobility and vehicles that no longer have a driver at the wheel. Because, after all, the vehicles will no longer have a steering wheel.
06 Anyone wanting to experience the benefits of autonomous driving today should get a Navya. Not dissimilar to a golf cart, this self-driving vehicle is manufactured by a French-American company, has a maximum speed of 19 kilometers per hour and is ready to drive immediately. However, at the moment a Navya can only be operated in enclosed areas such as resorts or airports. Price per vehicle: 250,000 US dollars.