One Hundred Years of Self-Driving Car History



Old illustration: a family of four plays a board game while a futuristic electric car drives itself automatically (1957).



The first unmanned vehicles were ships. After centuries of struggle with winds and waves, ancient sailors invented devices that tamed these forces of nature and replaced man. These were simple but ingenious solutions, like the sheet-tiller system that is still used today.



To make it, it is enough to take a jib-sheet (the rope that controls the small headsail), pass it through the pulley and back through the entire deck. You can complete the system by tying a gum-tack bow to the tiller (a stick that changes the direction of the boat). Now when the gust begins to bring the boat to the wind, the jib will pull the rope over the pulley and move the tiller, steering the boat in the opposite direction again.



Tricks like these helped smart sailors reduce the fatigue of long shifts at the helm in the era of sail. Today you can use them to open a can of cold drink and enjoy the splash while your yacht blasts the crests of waves like a train on rails. However, although tiller were used to control the first cars, the old technology did not make the transition from sea to land. In 1891, when Benz introduced the steering wheel to the world, this question was finally settled.



In fact, on land, after replacing animals with machines, the task of automatic control has become more difficult. Motorization was a significant improvement over the muscle strength of traction animals, but the payback was the loss of "intelligence." Very often, riders and even cabbies fell asleep with reins in their hands. Their obedient animals either simply continued to move along the road, or stopped.



However, cars and trucks require the driver's attention every second. Their rapid growth in popularity, along with the increased risk associated with their speed and weight, has spawned many experimental automatic control schemes. A 1925 New York demonstration provided a glimpse into the future of self-driving cars while surprising and frightening the public at the same time. Driving down Broadway in front of thousands of spectators, the upbeat American Miracle car moved, according to the New York Times , "as if the hand of a ghost were driving."



image


American Wonder, 1925



In the 1920s, motorized transport took tens of thousands of lives annually, and the death rate from this cause was 18 times higher than today. New technology promised to make city streets safe again. However, these hopes were soon dashed - first on 62nd Street, and then on Columbus Circle, the operators of the futuristic car lost control, after which the "miracle" collided with another car.



Despite this oversight, the automotive industry continued to dream of remote control cars. At the 1939 World's Fair, General Motors' Futurama booth displayed a huge motorized diorama of an American city. Unconstrained highways were filled with automated cars, trucks, and buses crossing busy areas of graceful skyscrapers. The diorama even had a "traffic control tower", in which, according to the designers of the city of the future, dispatchers had to radio control the movement of tens of thousands of cars. By the 1950s, guide wires embedded in the road surface had taken the place of a promising technology for remote control of transport. It's funnythat the first successful demonstration of such a system was first implemented in the 1950s by RCA (Radio Corporation of America).



These first prototypes showed the technical feasibility of automated driving, but due to the high cost and lack of demand, neither radio control nor guide wires gained popularity. According to calculations, the cost of the highway for controlled transport was as much as $ 200,000 per mile. If fully implemented, this "upgrade" of roads would add an additional 40 percent to the cost of building the US Interstate Highway System, already the largest public works project in American history. In the meantime, despite the dangers and hardships of long or night journeys, car manufacturers were still on the rise of consumer enthusiasm for cars. They focused on making powerful new cars that would be a pleasure to drive.



In the dreams of that time period, the future of automated driving was based on external control. By the 1960s, however, the focus shifted to using new computer technology to build cars that could actually drive themselves without outside help. At Stanford University, researchers pioneered robots that used cameras to track the road and computers to navigate. In carefully controlled experiments, these early droids moved along white lines and avoided obstacles in their path.





Stanford Cart, 1961



Autonomous Traffic was briefly laboratory projects. As processors and image processing techniques improved, by the late 1970s, engineers at the Mechanical Laboratory of Tsukuba University were able to test the world's first self-propelled passenger vehicle on the roads of Japan. Traveling at speeds of up to 20 miles per hour, these first automated vehicles used two video cameras to visually recognize road markings. In the 1980s, action shifted to Europe, where Professor Ernst Dieckmanns of the University of the West German Armed Forces augmented the Mercedes-Benz van with self-made autonomous driving devices, thus beginning a decade-long collaboration with automotive giant Daimler.





The autonomous machine of the University of Tsukuba laboratory, 1977



Finally, it was the turn of the United States - in the 1990s, Carnegie Mellon University took over the leadership. Autonomous car competitions were spreading around the world, software was being actively improved, and computers were getting faster, opening up new possibilities. By the end of the decade, the first automated trips across the country were recorded in the United States, Germany and Japan.



However, the most active period of development of unmanned vehicles (UAVs) has not yet begun. In the early 2000s, the Pentagon watched the development of this technology with increasing interest. In 2004, 2005, and 2007, the Defense Advanced Research Projects Agency, the most independent offshoot of the US Department of Defense and funding research, organized a series of open competitions to focus the work of individual research groups and strengthen ties with the defense and auto industries. These "Grand Challenges" offered millions of dollars in prize money and priceless prestige and attracted dozens of teams from academia and industry.



Testing their best hardware and software, the competitors watched from afar as their UUVs attempted to navigate open spaces and more urban landscapes at an abandoned military base. The 2004 competition ended with no winners - none of the competitors were able to reach the finish line. However, a year later, Stanford University's winning car brought him a $ 2 million prize.



The DARPA competition has accelerated the development of unmanned vehicles. Stanford won first place for its advanced use of machine learning in video processing, one of the AI ​​programming techniques. But more importantly, the competition drew attention to the possibilities of emerging technology. No one was surprised by the growing interest of the military in the UAV. However, their potential application in the civilian sphere sparked a great deal of thought. For the first time, people felt that the practical commercial use of autonomous movement technologies was already achievable.



It was a call to action for the auto industry, but not everyone heard it. Most companies were overly preoccupied with the 2007-2008 financial crisis and the ensuing financial downturn. US auto makers have experienced particular difficulties in using the capabilities of the BPA, because this would require significant investments to transfer technology from laboratories to the market. The car manufacturers went bankrupt or the federal government came to their aid. Meanwhile, Silicon Valley moved forward. By 2009, the head of Stanford University's DARPA-winning team, Sebastian Tran, was already managing the self-driving car project at Google. The search giant has made a big bet on Android, its highly successful mobile phone operating system. It seemedthat passenger cars could become an important new computing platform. Can Google Contribute to the Future of Automotive Software? The stake seemed right, fueled by a long-term interest in BPA from CEO and co-founder Larry Page.





Google's new BPA showcases at Google X in Mountain View, California on May 13, 2015.



It took the world several years to comprehend Google's move, after which explosive activity began, not only in the automotive business, but also in the computer industry, as well as in the taxi industry. All of a sudden, every major automaker, every company involved in the road transport of people, and rival cloud software giants like Apple quickly mobilized their forces to develop self-driving cars. If their own projects did not produce convincing results, many companies simply acquired promising startups in order to get their hands on the necessary technologies. In just two years, 2016 and 2017, more than $ 80 billion in investments flooded into the field of BPA technologies.



The largest deal was Intel's panicky 2017 acquisition of leading computer vision developer, Israel's Mobileye, valued at $ 15 billion. As the flurry of such mergers and acquisitions unfolded, the network of partnerships and cross-holding companies linking car manufacturers to the tech sector became even tighter. Two of the largest consumer industries, the computer and automotive, saw their future in each other. But they couldn't decide if they wanted to move together or take over competitors.



In 2018, hard work and massive funding paid off. In December, Waymo, a subsidiary of Google, quietly rolled out the world's first self-driving taxi service in Chandler, Arizona. More than 40 years after the first test drive in Tsukuba and nearly a decade after Tran's hiring, the company began accepting orders for unmanned rides in the Phoenix suburbs. The tech giant was reported to have allocated more than $ 10 billion to build its BPA empire. It seemed that the long and painful process of the birth of the BPA was finally over.



"There is hardly any problem for horse-drawn transport that could not be solved as well, and perhaps even better, with the help of cars," wrote the New York TimesJanuary 12, 1903. At that moment, the first major auto show opened at Madison Square Garden, then located on 26th Street and Madison Avenue. A century later, the Times similarly shared its enthusiasm, this time observing the engineering wonders of the drone age. In 2018, columnist David Leonhardt wrote, "On my fourth day of driving a semi-automatic car, I was ready to take the leap into the future."



This newspaper was not alone in its enthusiasm. Like the emergence of cars themselves, the development of UUVs has spawned bold predictions about the benefits the new technology can bring to people and society. But what does this future promise us?



First, according to their researchers, self-driving technology can eliminate almost all deaths caused by cars. An estimated 60 million people died in motor vehicle accidents in the twentieth century. This is more than all the military and civilian losses during the Second World War. But even as cars are becoming much safer, deaths continue to occur as motorcycling is spreading to new countries where experienced drivers and traffic regulations are in short supply. Amid the auto boom in China and India, more than 1.4 million road deaths occur worldwide each year. The vast majority of accidents could be prevented with self-driving technology, advocates argue.



Secondly, BPA will solve the problem of traffic jams. The economic losses from congested roads are enormous; the location tracking devices found in modern mobile phones make them much easier to assess today. Using the sheer volume of travel records left by telephones, news agency Inrix has calculated that in the United States alone, drivers spent more than $ 305 billion a year in traffic jams, or about $ 1,500 per driver. The argument in favor of the FUU is that the vehicles driven by the software can maintain a shorter distance between each other on the highway due to their faster braking reflexes. In addition, FUAs can also reduce congestion simply by sparsely dispersed communities.When UAV passengers can use their commuting time to work or rest instead of tracking the road, longer trips to less congested areas will become less of a burden.



Third, technology proponents hope that everyone benefits from BPA. In the twentieth century, automobiles increased the mobility of hundreds of millions of people, but the success of the automobile dispersed the population and led to an outflow of funds for mass transportation, so many faced new obstacles to free movement. In the United States alone, more than 25 million people have disabilities that impair their ability to move - about one-sixth of the workforce. It is believed that UAVs will not only provide car travel opportunities for those who are physically unable to drive, but will also open up new opportunities for the very elderly, very young and those who cannot afford their own car. If people with disabilities step out of the way of life and join the workforce, if senior citizens have convenient access to health care,and if children have a wider choice of educational and developmental services, the social and economic benefits of this will be enormous.



You ask: when will this utopia come? Today, BPA is still a novelty for us. Despite all the difficulties, dangers and worries of driving, we are still the most cost-effective "technology" for this task. If we have surpassed even our wildest dreams by the time you read this article, there will still be fewer than one million self-driving cars on highways, streets and roads around the world. However, over the next decade, the number of BPA is doomed to rapid growth. By 2030, the total number of smart cars, trucks and buses could rise to tens of millions. They will share the road with about two billion people-driven cars and trucks (give or take a few hundred million). Seem to be,that even then the number of BPA will be approximately equal to the rounding error of the total weight of the vehicles. But the revolution will come suddenly, with surgical precision and crushing power. As the author of cyberpunk books William Gibson once said: “The future has already arrived. It's just that it's not evenly distributed yet. "





Uber, Uber Advanced Technologies Center 13 2016 , .



The first changes that we will notice will occur in the taxi industry. Most market analysts agree that all taxis in industrialized countries will be automated by 2030. In the USA, this amounts to 300 thousand cars. If you add up all of Uber and Lyft, this adds up to about 1 million. Driving from airports and resorts to our beloved suburbs, driverless cars could become the face of automation for a generation, as well as the start of unmanned travel for billions of passengers every year. The advent of driverless taxis has the potential to radically change how consumers perceive cars. When computerized chauffeurs are just a tap away, people may decide to give up car ownership altogether. If this shift in consciousness becomes massive,then moving the same number of people will require a much smaller number of cars than there are now private cars.



But this idyll may not come. Automation will also make private cars more useful, and software will drastically reduce the burden of owning them. Think about this for a minute. Automated cars will not only be capable of driving - they will be able to park themselves, drive to the garage for fuel and repairs, and also pay their own insurance bills (with your money, of course). It is likely that we will simply replace our “stupid” cars with smart ones and continue to drive as before.



In the long run, we are likely to get something mixed. By 2040, even if general-use UVVs take over the initiative and new passenger car sales drop 50 percent (a really big change), automakers will still be producing roughly 30 million self-propelled cars around the world. Half of them will end up in China, another quarter in the US, and the rest will be distributed across the EU, Japan and new markets. But even if the passenger car business diminishes, the business of using cars, as well as vans, scooters and anything that can move, will grow. What remains of today's $ 2 trillion international auto market will be swallowed up by the much larger "personal transportation" market.which is projected to grow from $ 7 to $ 10 trillion by mid-century - roughly the size of the entire modern EU economy. Waymo alone wants to capture an annual share of 1.7 trillion by 2030. However, Uber, Amazon and Alibaba, not to mention Ford, GM and VW, will not leave this new market without a fight. They have their own projects in the future drone service businesses.



Therefore, although the drone revolution began with a thin trickle, very soon this slow trickle will develop into a stormy stream. By about 2050, most human-driven cars will be gone. They will be replaced by a smaller, smarter fleet of autonomous vehicles of all shapes and sizes. Some of the vehicles will be private, others in public use. Some will be able to move one person, others will accept a hundred or more. Many will not carry anyone at all, but will occupy themselves with transporting the incessant flow of food that has arisen thanks to the triumph of online shopping. Some will help us simply by observing our urban world or regulating traffic. Overall, it can be said that a diverse fleet of UAVs will travel far more miles than vehicles today.



I would like to say that the unmanned revolution will repeat the history of the 20th century with cars, only on a larger, computer-controlled scale. But nothing in our past can prepare us for what lies ahead. The pace of the speeding change will overwhelm us. In the United States, full motorization took about 60 years - from about 1920, when large numbers of cars began to appear in cities, until 1980, when cities began to choke on them. The next 40 years, from 1980 to 2020, were a period of saturation.



The average number of hours commuting to work in traffic has nearly tripled, while economic losses from congestion have increased tenfold to $ 166 billion annually. We spent most of this time looking for ways to reduce car use and investing in alternatives.



However, automation can have its effect in as little as 20-30 years - within one generation. If our history with cars has taught us anything, it is that the autonomous revolution will not be what we imagine it to be.



All Articles