Driving for a living
Nico Rosberg introduces us to self-driving cars, considered by many to be the inevitable future of transportation.
Ok, I admit it – when I found out that I was going to be featured in Quattroruote with a big sign saying “So long, driver,” I felt a bit unsettled, even somewhat offended. “Why,” I wondered, were they asking someone who had made a living from driving racing cars to come endorse cars that drive themselves? But after getting over the initial hit to my pride, I decided to do some research on the so-called ‘autonomous’ cars – and on the projects that my colleagues at Mercedes have been doing in this field. Eventually, I grasped the basic fact that the objective of this new automotive know-how has a precise objective – namely, that of safety. I leave it to the engineers from Stuttgart to judge the potential impact that self-driving may have on general transport. Though not an expert, I can nonetheless sense the enormous social advantages that robots will present, if and when (presumably a few decades from now) they become a reality. The result will be a world without accidents and, therefore, without human tragedies or collectively burdening costs.
We’ll travel on free-flowing streets because computers will be able to predict traffic jams. We’ll consume less fuel, liberating ourselves from the shackles that tie us to fuel-producing countries. We’ll have more time to spend on our interests or on work, while an invisible chauffeur leads us to our destinations – we’ll be able to watch TV, surf the web, or play with our children. Silent guardian angels will watch over us while hidden control units exchange data and locations – the only traffic will be that on Information technology highways on which electronic installations will be a Bablyon of electric bits. Eventually it occurs to me that if we accept, without problems, the fact that the planes on which we fly are monitored by control towers thousands and thousands of kilometres away and that they fly 99% of time by relying exclusively on different aviation instruments – then why should the same not happen with cars?
At this point, the question that I ask myself is this – will all of this mean the disappearance of the driver, and, ultimately, of motor sport in which the man behind the wheel is the true centre? I don’t think this will be the case. Firstly, as a driver I am aware that our professional effort is contingent on the progress of the auto world as a whole, as illustrated by the innumerable solutions migrated from the racing tracks to everyday cars. Turbocharged engines, direct injection, systems of traction control, sequential gearboxes, regenerative braking, are just some of the ideas born in the racing world that now avaliable in road cars.
Racing will, therefore, remain an essential laboratory for standard production vehicles. And the same is true in terms of the human element. I really don’t think that one day we’ll have mono-seaters that run by themselves – partly because, if that’s the case, I would be out of a job and also because, philosophically speaking, racing is nothing but a metaphor for life rather than a confrontation between mechanics. What’s important is that we conduct scientific research that’s compatible with the human touch – because passion and emotion will always travel on four wheels.
The Future Is Self-Driven - Autonomous Vehicles
According to many automakers and analysts, robot-cars are at least a few decades away. There are advantages and limitations to a technology that promises to revolutionize the world’s streets.
The first impact, in the true sense of the word, is terrifying. The Mercedes S-Class advances towards the dummy placed in the centre of the Sim City track in Sindelfingen. The engineer, sitting in the passenger seat, politely tells us not to worry: “The car takes care of everything, you can go ahead and cover your eyes with your hands.” We obey, but not without first crossing our fingers the way you do when you’re watching a horror movie as a kid. We’re travelling at 50km/h, and the dummy is approaching. Suddenly, the car screeches to a halt. The Pre-safe braking system has identifies the ‘pedestrian,’ and it has read its height and position, following which it’s told the central unit to halt the car. The problem is that, as humans, we are notoriously imperfect – or at least more imperfect than the radar that scans the landscape in front of us, always adjusting to the car’s pace. At the first sign of slowing down, the instinct is to hit the brake pedal. The problem is that we’ve run out of room. Bam. The dummy is hit square on and flung a good distance back. “You’ve killed it,” says the co-pilot. “By braking, you told the system that you wanted to regain control. It complied, but by then it was too late to stop the car. In other words, it was your fault.” Thanks, all I needed was an additional dose of guilt. From then on things gain greater fluidity. You have to learn to trust the system entirely, to trust the sensors and the control units that send the signals. You need to suppress all natural impulses, to forget years and years of conditioned reflexes. ‘She’ is the one driving, the S 500, a travelling laboratory full of all the active and passive systems in Mercedes’ safety know-how – Pre-safe, Night vision, Distronic, BAS plus, Active lane assist, Collision prevention, along with the usual ‘banalities’ such as ESP, ABS and other acronyms. Don’t see the pedestrian? She stops. They cut you off on the street? She stops. You get distracted on your phone and slide slowly towards the emergency lane? She brings you back in line. You go to overtake a car and there’s another one coming in the other direction? She rejects the manoeuvre. We travel a considerable distance, and the sedan shows not a single hesitation or a second-thought. Eventually, just to exaggerate, we take our hands away from the steering wheel – it starts to manage on its own, using the lines on the street as points of reference. After a few seconds, the alarm beeps – this means enough horsing around, take control of the steering wheel. This S 500 Intelligent Drive, in fact, is not the car that drives itself, or, in popular jargon, the ‘autonomous car.’ Or, at least, not yet. As the head of the Driver assistance & active systems project Jochen Hermann says, “For the moment, we’ve parked our self-drive car in one of our laboratories. We need the S Intelligent Drive to collect all the information that we can get on the functioning of electronic systems, and to test the interaction between man and machine. Jochen makes one thing needs very clear, “On the S-Class that you drove today, the technology necessary to make the car drive without human intervention is already there. The autonomous car is technically feasible – now the task is to create a surrounding environment, in other words to adjust the regulations, to resolve the issue of responsibility in the case of an accident as well as building a new infrastructure.” How long will it take? “Better not to get ahead of ourselves. But a realistic time frame would be within the next 15 to 20 years.”
Mercedes is not the only automaker investing money and resources on self-drive cars. Nissan, for example, is so sure of itself that it has set 2020 as the precise launch date for its first autonomous cars. Similar efforts are being made by BMW, Audi, Ford, Honda, Toyota, GM, and even ‘the outsider’ Google – whose driverless Prius has become omnipresent on the streets of California, Florida, and Nevada. Now, such a race between automakers so different from one another is no coincidence. Everyone believes that the driverless car will be an inevitable part of the future – one that’s not as far off as everyone thinks.
Morgan Stanley agrees – it’s convinced that the robot-car will bring a revolution in the field of transport. In order to corroborate such a claim, it has delineated a precise picture of the potential consequences of such a historical shift. The first phase will be the already existing ‘passive’ phase; the second phase will see the partial substitution of the driver between 2015 and 2019, and, finally, between 2019 and 2022, the final phase of total autonomy. Different consultancies have hypothesized that the arrival of self-driven cars could mean a total socio-economic saving of 1.3 billion dollars – every year. That amounts to 8% of America’s GDP. It’s in this mind-boggling figure that the scope lies for an innovation that seemed destined to remain in cinematic fantasies such as Supercar or I, Robot (see the box on pg. 102). Firstly, these cars will have repercussions on the number of accidents, even if Hermann feels the need to clarify. “Autonomous cars are not the means to secure an accident-free world – if anything, they are the consequence. And the more of them that begin to circulate, the more drastic the drop in fatalities will be.” A study done by the ENO centre for transportation gives exact data – if 90% of the cars in circulation in the US were self-driven, the result will be that of 21,700 fewer deaths a year. Public expenditure on hospitals will also reduce as a result. There would also be more direct advantages. Just in the United States, according to Morgan Stanley, where the global consumption is of 143 billion dollars of fuel a year at a cost of 500 billion dollars, an autonomous car, by avoiding traffic and maintaining a consistent pace can increase range by 30% - everything else being equal. And that’s not all. The capacity to predict the best route will also have an impact on social productivity – both because of the journey time saved, and the fact that while the virtual driver takes care of the driving, we’ll be able to occupy ourselves with other things (consultants, of course, just assume that one will necessarily choose to work in the car…).
The ethical issue
So, no flaws at all then? Well, not really. The real problem with these autonomous cars is not so much their technology, or the cost for the client (Mercedes predicts a price increase of just a few thousand Euros), but the adjustment of the social apparatus. Intelligent cars need intelligent streets, which will no longer be inert objects, but dynamic systems that will manage the interaction between cars, asphalt and signs. And drivers will no longer be simple users, but generators of information for an integrated management system. Lastly, a great unknown needs to be resolved – the role of the driver. In an automated world, the concept of free will and, more generally, the ethical approach towards civil cohabitation are destined to change. The human element – with its errors, unpredictability, but also its intelligence – could become an optional factor rather than a determining one. Pure sophistry? In reality, there are real practical implications to consider. Above all, in the case of an accident who will be responsible? Brad Templeton, the autonomous car guru (he was a Google consultant), reasons: “No one can hope to reach absolute perfection – despite autonomous cars, there will still be accidents, and I think that there always will be. I’m convinced, however, that the steering wheel will never disappear – autonomous driving will have a switch allowing the driver to assume control if and when he or she desires to.” In our opinion, there is an essential missing variable in all of this, which is that, for millions of people, driving is still a pleasure. Not to mention that the car has always been an instrument with which to distinguish oneself individually and socially. These are intangible variables, therefore not on the discussion table. But they have always been part and parcel of the car’s success and popularity, beyond their simple utilitarian usage.
Travelling the world alone
The autonomous version of the Mercedes S-Class during one of its test drives. By law, the driver’s presence behind the wheel is still obligatory
PROHIBITIVELY EXPENSIVE RADAR
The base is that of a Prius, but the autonomous driving system is not Toyota’s, but Google’s. The project is directed by Sebastian Thrun, who was the co-inventor of Google Street View and is now the head of the Stanford artificial intelligence laboratory. It was his Californian group, which won the Darpa Grand Challenge in 2005 (that prototype had the base of VW Touareg) and the prize of 2 million dollars awarded by the US Defence Department. As with the drone planes, the military has been following the development in self-driving vehicles since the start, and at the beginning were even funding the research. The Google Prius features equipment that today costs well over 100,000 Euros. The laser radar (called Lidar) is particularly expensive, costing over 60,000 Euros. Too much! Moreover, this cumbersome sensor emitting 64 laser rays needs to be installed on the roof, which defaces the aesthetics of the car. In the present state, the Google system needs the Lidar to generate an extremely detailed 3D map projecting the Prius’ surroundings. Other sensors on board include a camera on the windshield and two more radars, one at the front and another at the side. Google already has at least a dozen autonomous cars and at the end of 2012 it announced that it has covered over 500,000 kilometres with them without a single accident. This was due to the fact that California, Florida, and Nevada have approved laws that allow their use on public roads.
THE REAL DIFFICULTY IS IN THE CITY
The Nissan Autonomous drive system was introduced to investors and journalists in Irvine, California, at the end of August 2013. Two prototypes based on Electric Leaf technology were revealed – one equipped with the systems necessary for highway driving, and the other with systems (far more sophisticated ones) that are indispensable in the city. The designers are now working step-by-step to combine the two technologies into a device that will produce a commercial Nissan autonomous vehicle by 2020. The Leaf designed for cities has five lasers and five cameras that monitor what happens around the car – since here, unlike on the highway, there are stoplights, intersections, cyclists, and pedestrians, so the work done by the electronics is exceptionally complex. That’s why Nissan is planning to build a test track for this exclusive purpose.
THE SWEDISH TRAIN
Volvo’s approach is gradual and pragmatic. Months ago, it concluded its three-year project called Sartre, dedicated to cars capable of following the traces of another car (usually a trailer truck). All of this thanks to a thick exchange of information, transmitted via antennas and specific communication protocols. This allows you to drive on the highway like a platoon, in groups of up to eight cars. Contrary to what happens to cars driven by humans, the braking of all the vehicles that precede the group happens with a maximum delay of 0.02 seconds – so, the last one in line doesn’t risk crashing into the car ahead. The system, however, is not appropriate for the city. The most interesting aspect of Sartre is that it utilizes hardware that’s already on board of the Volvo’s currently on sale – cameras, a distance-laser sensor (perfect for short-range), a radar (indispensible when trying to identify distant obstacles). The only unique elements are the software and the communication system. In other words, the technology is cheap and far simpler than the technology which allows cars to drive in truly autonomous fashion.
FROM PIKES PEAK TO THE PARKING AREA
In 2010, an Audi TTS driverlessly travelled up the Pikes Peak mountain road in 27 minutes – not bad keeping in mind that a pro driver would have taken around 17. The car had been developed by the Electronic research lab of Palo Alto. The director said, “We’re not trying to replace the driver, we just want to develop the best drive-assistance system possible.” Audi says it’s ready, beginning in 2015, to sell cars capable of controlling the accelerator on their own, along with the brakes and the steering at low speeds in order to free the driver to occupy himself with whatever else they want to do while waiting for traffic jams to break up. Then, sometime before 2020, the automaker will propose a technology capable of making a car self-park itself via a command emitted by a smartphone app.
YES TO ELECTRONICS, BUT ONLY PART-TIME
BMW and Continental collaborate in order to develop the electronic co-pilot, which is conceptually different from the driverless car. In fact, the Munich-based automaker’s main worry (and that of others…) is that if something were to happen which the electronics were unable to control, the driver shouldn’t be entirely estranged from the driving. For example, the driver shouldn’t fall asleep. BMW’s approach, therefore, will be to offer 99% autonomous driving but never 100% - part-time rest to allow phone calls, e-mail and news reading through the Connected drive that’s connected to the internet. This idea appeals strongly to the wealthy business class. BMW’s objectives are clearly defined – partly-autonomous cars in 2016, automated in 2020, and 99% self-driven by 2025.
Jochen Hermann leads the Mercedes Driver assistance program. In his opinion, the technology needed to make self-driving cars already exists, but the market still needs to be ‘built’
LIMITED TO OBSTACLE PREVENTION
Autonomous driving? No thanks, says Toyota. It prefers to invest on devices that assist the driver rather than substituting him. One of these is the ‘safe-pedestrian’ system, which, once the obstacle has been identified, automatically activates the brakes. And if this were not enough, the car tries to dodge the obstacle by operating the steering. The steering wheel angle can reach 90 degrees, but the correction is activated only in the presence of a clearly delineated lane. The Japanese group is working on a highway driving assistance program (automated highway assist) that adds lane control to the habitual cruise control functions.
We’ve already (not) driven it
Above, the article from the March 2013 edition of Quattroruote when we were offered the chance to test an experimental model of the E-Class Intelligent Drive
The S500 Intelligent Drive screeches to a halt automatically in front of the dummy, while the driver removes his hands from the steering wheel and looks elsewhere
PLANNED EVEN BEFORE 2020
We’re at the Honda research and development centre in Tochigi, Japan. On board the Accord, there is a driver, but he doesn’t touch the steering or the pedals. The car uses radars and cameras to evaluate the width of the street and to ‘see’ pedestrians or any other obstacles. Before they even have the chance to spring from behind a corner, she spots them – so long as they have a smart phone in their pocket and an activated wi-fi. Okay, so the system is still in its experimental phase. It’s anyhow impressive, as a passenger, to see the steering wheel turning on its own and the car accelerating or braking without receiving any input. At Honda, according to Toshihito Nonaka, vice-president of Honda’s R&D department, autonomous driving will be available very soon – even before the 2020 mark used by most.
Film directors that saw into the future
Dal Maggiolino tutto matto fino a Spielberg, il mondo della celluloide aveva già previsto l’avvento dei veicoli in grado di viaggiare senza pilota
From the crazy VW Bug to Spielberg, the cinema world had already foreseen the arrival of driverless cars.
In the beginning, it was Herbie, the VW bug. From that moment, the cinema and television world (not to mention cartoons) have always been fascinated by cars that are able to drive themselves. The rage began with David Hasselhoff, who commanded his ‘Kitt’ from a distance (the TV Show, which first aired in 1984, was called Knight Rider) – then there was an avalanche with unexpected insights into the automotive future. In the 2004 film I,Robot, set in the year 2035, Will Smith is seen in a driverless Audi. In Spielberg’s 2002 Minority Report, Tom Cruise drives a futuristic missile-like car without driving it.
Above: Herbie was the protagonist of six films (the last one was in 2005); the TV series ‘Kitt’ was based on a Pontiac Trans Am; the Audi in I, Robot was called RSQ and evoked the concept Nuvolari. On the streets of Minority Report, Tom Cruise travelled in a voice-commanded capsule.
© Riproduzione riservata