Autonomous Cars: New Age of the Modern Automobile
The future has arrived and it’s being driven in by an autonomous car.
Not long ago I chatted with a relative about modern invention and future innovation. Having done no research to support my hypothesis, I propounded the possibility of driverless automobiles and a technology, certainly already in place, that would make driving foolproof and safer and render human steering near-obsolete.
Perhaps the idea is Jetsonian and what people fifty years ago may have anticipated with the dawn of the millennium. I would have been humbled to learn that my notions were close to the truth; but I couldn’t have expected them be real, for they were dead-accurate.
The age of the robotic car is here and will not wait another fifty years to hit the streets.
The History of Autonomous Cars
Jared Diamond, in his Pulitzer winning Guns, Germs, and Steel, expounds on the common expression ‘Necessity is the mother of invention’. He explains that often it is not necessity that creates a demand for innovation, but rather innovation that makes a place for common use. So to ask the question “Why an autonomous car?” may be as futile has having asked why a computer or the internet before we ever knew to need them.
The robotic car was envisioned very early in the 20th century, and several countries have endeavored to create a fully autonomous vehicle. It wasn’t until the 80's—after a few simple robotic machines in the 60's and 70's—that the first driverless automobiles were built and put to the test.
Ernst Dickmann of Bundeswehr University Munich (GER), who has been compared to the Wright Brothers in aviation, was the first to build and test driverless cars in the 80s. His vehicles reached 55 mph on empty streets. He was the first to demonstrate the idea’s real possibility; and since then an international race to build the first fully autonomous vehicle has ensued.
DARPA and Autonomous Vehicle Research
Germany has led the pursuit for driverless cars and remains a leader in its research. The United States, however, has also joined the race and in a major way.
In the 80's the Pentagon’s Defense Advanced Research Projects Agency (DARPA), the nation’s foremost research group, funded the Autonomous Land Vehicle, which was the first of its kind to use new technology like laser radar and computer vision, still key components in today’s research. With the progressive success of the vehicle, the agency created a long distance competition for driverless cars throughout the world, called the Grand Challenge (2004). In fact, the U.S. Congress authorized DARPA to offer prize money—$1 million—to the winner of this challenge.
Only a year later and since that first competition—with good international participation, the continued success of the competitions, and new and different aspects of the challenges—the prize money has doubled and research is greater than ever.
The military’s purpose for robotic vehicles did not intend to spur interest in the mass production of such cars; it only meant to facilitate robotic development with the goal of making a third of its ground forces autonomous by 2015.
Advantages of Autonomous Cars
The benefits of autonomous research, however, reach far beyond the military’s hopes. Not only is the idea an awesome invention, it comes with several major benefits to society.
The most poignant is safer, automated driving that would reduce traffic fatalities. The World Health Organization states that about 1.3 million deaths worldwide—43,000 in the U.S.—result each year from roadway incidents, and most of them are attributed to some type of human error. An autonomous car though doesn’t get distracted, sleepy, or drunk, has 360-degree vision, and is far faster to react than a human.
Scientists also say that robotic cars would effectively double the capacity of roadways by allowing cars to drive safely while close together. (More on the technology that makes this possible later.) Would this mean more traffic? Perhaps, but traffic would now be intelligently controlled and far less prone to accidents.
Passengers would be most directly affected and not just because people may no longer drive. Suddenly restraints fall to the ground with autonomous vehicles: If cars drive themselves passenger's ages no longer matter and neither do their physical or mental ability to operate a vehicle. Disabilities disappear as well to some degree. This is a unique benefit for the elderly who sometimes need to get around but may be dependent on others.
Even more, passengers wouldn’t necessarily be required at all—the car could drive itself to where it needs to be. Some airplanes are already capable of this. The notion figures prominently in the concept of the shared car that could be summoned to pick up, say, several co-workers at one time, drive them to their work, and then service others after the workers have exited—like a taxicab. The concept conceivably reduces the number of cars on the road entirely.
This all leads to unintended benefits as well. If there are fewer cars there will be less need for parking, freeing up valuable acreage. There will also be less need for policing. Then, there will be less need for gasoline, a major sustainability topic. And with less crashes cars can be built lighter thereby reducing the need for raw materials and, once again, fuel consumption.
The Technology Behind Driverless Cars
Autonomous cars require five components of technology to operate: radar, optics, laser radar, GPS, and processors. Traditional radar is used to detect objects in the vehicle’s path more than 100 meters (328 feet) away. Video cameras are used to see road markings and traffic signals. Laser radar scans the car’s surroundings in a 100 meter radius. This information is sent to the processors (as many as 17) that make sense of the data by generating a detailed map of surroundings for use to avoid objects. The GPS keeps the car on its route with an accuracy of 30 centimeters.
Such technology, for instance, allows a 2-second human reaction and braking sequence to be trimmed to .3 seconds.
Infrastructure for Robotic Cars
The biggest hindrance to seeing autonomous cars in the next few years are three: lack of infrastructure, legislation, and unknown cost of manufacture.
It is no secret that the nation’s infrastructure already needs overhauling; and this is one reason why fully autonomous cars will gradually emerge in the culture. The systems that have been imagined for the new technology right now are only considered for use within a limited area or stretch of highway, not across entire cities or regions.
But these systems—and there are three—are where the ‘cool factor’ about robotic cars comes to the fore. One is a dual-mode transit system in which a person’s private vehicle could be serviced with the benefits of a monorail system. The car would be designed with the ability to dock into a public monorail system and, being managed from a central mainframe, become part of a driverless transport system. This is not unlike the world of George Jetson.
The automated highway system would again allow a private vehicle to link to an automated road system. The idea here though is to use the technology to keep a vehicle centered in its lane and allow it to communicate with other vehicles to manage traffic.
The most interesting system though is the free-range system. It would allow vehicles to locate themselves in the system by GPS and plan their own routes. The main system would only need to oversee the operation and direct traffic as needed.
Driver Autonomy v Driver Assistance
Companies like GM, Volvo, and Volkswagen are already testing autonomous cars; but not all seek the full autonomy of their vehicles. BMW, which is also testing robotic cars, says that a fully autonomous car does not comport with its client base. And some research is indeed headed the way of autonomy but is not completely aimed at that goal.
In fact, many of the driver-assistance mechanisms added to cars in recent years is but the gradual move toward the fully robotic car—on the low end with features like anti-lock braking and stability control to the much higher end with lane departure and eye monitoring systems and cars that park themselves.
This is often the way technology happens to us—gradually. It never overtakes us at once but prepares the way for its arrival.
Google's Driverless Car
- Sebastian Thrun on the Google Driverless Car
Stanford professor Sebastian Thrun leads Google's research on the driverless car. Here he presents a 6-minute discourse and fascinating footage of the research.
When Will Autonomous Cars Hit the Streets?
General Motors has stated that it will begin testing driverless cars by 2015 with the possibility of having them on roads by 2018. It is the earliest prediction I have found, the latest being 2056. So most of us will witness the advent of the autonomous car at least in its crudest form and not as the common convenience our grand- and great-grandchildren will experience.
Google is surprisingly a premier researcher of the driverless car. Already its fleet of tech-laden Toyota Prii (the plural of Prius) and one Audi has successfully logged 160,000 miles without driver intervention; and the company has set a new challenge of one million miles. Not only were the cars’ artificial intelligence able to produce detailed maps of the environment and know all its speed restrictions, they were also capable of detecting out-of-the-ordinary cautions and being programmed for different driving personalities.
It is not clear yet how Google will profit with its research and advances, but in one way it has already made a moon-landing impact.
Legislation: The Last Hurdle
The problem most often encountered with new technology is outmoded law—and whether it can be called outmoded is questionable because often the real issue is technology simply being well ahead of the law as we have witnessed often in the cyber world. Legislation seems to only focus on the present and the precedent, but never the possibility.
A formidable legal challenge to the autonomous car is federal safety standards. Only until driverless cars can be proven exceptionally safe—far above requirements for humans—will they be approved for the roads. And this inevitably leads to other questions: What if there was an accident somehow caused by a driverless car? Who would be to blame—the operator or the software?
Google has been at the forefront of this legislation, however, and has been able to cut through much of the hassle with answers from its research. For instance, the law requires that a person to be in control of a vehicle at all times. Yet never once were Google’s experimental cars out of the control of a person. Attendants were able to assume control of the vehicle at any time, even to override error if necessary.
Through such examination and careful argument, as well as a hidden agenda, Google has lobbied two bills in Nevada that have blossomed gloriously. In June 2011 Nevada became the first jurisdiction in the world where driverless vehicles can be operated on public roads. This doesn’t mean that robotic cars will be zooming around anytime soon, but it’s a major stepping-stone.
The Future Is Now
In the next 100 years our roadways may be completely different from what they are now. Very few of us will be driving—free to text, sleep, and work—and more of us will own fewer cars and the chores of maintaining them. Just as the Ford Model T was a social phenomenon, so will the fully robotic car be an engineering marvel that will redefine how we measure cultural convenience.
The Jetson—wouldn’t that be a cool name for the first fully robotic car? I doubt that our common modes of transport will ever become what we see in sci-fi films and cartoons (maybe by the next millennium). And neither will we be able to fold our cars up in our briefcases like George Jetson does in the cartoon intro. We’ll leave that to the computer, which is headed that way. For now we’ll anticipate and be wildly excited about a fully autonomous car.