About the Project

The future has always been hard to predict, but as technology evolves ever more quickly, it can now be just as hard to imagine. To understand the next wave of innovations—such as nanotechnology in our bodies, computers that can see like humans, and the Internet of Things in the age of scarcity—Qualcomm collaborated with Atlantic Re:think, The Atlantic’s creative marketing group, to explore the emerging edge of technology through art.

Part 5

THE SPACE
AHEAD

Thanks to breakthroughs in processing power, connectivity, and artificial intelligence, self-driving cars are no longer the stuff of dreams. In the coming age of autonomous electric cars and hybrids, the aged and disabled will be freed from their confinement, carbon emissions will be slashed, and most highway deaths will be eliminated. That’s why they’re called “smart cars”.

select

Ushering In
The Age Of
The Data-Driven Car

In another case of digital disruption, the global automotive industry is in the midst of a revolutionary paradigm shift, turning away from its century-old dedication to brawn—power, speed, and body design—to a focus on brains.

THE SPACE AHEAD Visualized by Michael Murphy

As every inventor knows, the mother of innovation is failure. In 1885, Karl Benz started his first automobile for a public demonstration and crashed into a wall. In 2004, the U.S. government put up $1 million in a “Grand Challenge” to deliver a self-driving car that could negotiate a relatively traffic-free 130-mile course through the Mojave Desert. Of the fifteen entries, the best made it only 7.4 miles. One flipped over before the start. Hours from the finish, all were broken down, disqualified, or withdrawn.

Everything depends on what comes next.

By 1888 Benz was selling his “Motorwagons”, and in 2005, at the second Grand Challenge, several teams’ vehicles finished the course, and the winner left $2 million richer.

That team was led by Stanford computer-science professor Sebastian Thrun, who began with a key insight: “Many of the people who participated in the race had a strong hardware focus,” Thrun recalled years later in an interview with Foreign Affairs, “so a lot of teams ended up building their own robots. Our calculus was that this was not about the strength of the robot or the design of the chassis…. We decided it was purely a matter of artificial intelligence. All we had to do was put a computer inside the car, give it the appropriate eyes and ears, and make it smart.”

vision

Michael Murphy, visualizes how technology and the automobile will reshape our world.

With that, Thrun and his team radically redefined the Great American Car, replacing its signal virtue of brawn—the growl of horsepower, the rush of 0-to-60 acceleration, the power and freedom of being “behind the wheel” out on the open road—with brains. After 2005, the automobile, like much else, would never be the same.

QUALCOMM TECH INSIGHT

“This is an incredibly exciting time in automotive, a lightning-speed inflection point: moving from a focus on the body of the car to the brain.” - Patrick Little, senior vice president and general manager, automotive, Qualcomm
More From Qualcomm

Thrun’s Stanford team won, not coincidentally, just as the study of artificial intelligence emerged from what was known in the field as a long “AI winter”. A string of defeats that had discouraged researchers in the field was decisively broken in the mid-2000s with critical advances in artificial neural networks. The development of “learning algorithms,” which mimic the synaptic connections between actual neurons, gave computers the ability not only to detect complex structures in huge datasets but also to learn from what they found.

This and later progress in “deep learning” sparked rapid advances in facial recognition. It increased the intelligence of drones, virtual assistants, speech-recognition and translation software, and recommendation algorithms that could tease out our tastes and preferences. One algorithm was recently shown to diagnose heart failure from MRIs as well as a cardiologist.

Such AI advances also spurred the R&D required to build a self-driving car that could negotiate not just a stretch of the Mojave Desert but also the dense complexity of urban streets, a world of railroad crossings, traffic lights, construction sites, emergency vehicles, bicyclists, pedestrians, the hand gestures of traffic cops—not to mention the behavior of other cars, trucks, and buses on the road, with drivers and without.

That complexity would have been impenetrable without an exponential increase in computing power (largely mobile-driven advances in graphics and digital-signal processors as well as CPUs), which made it possible to deal with enormous quantities of data. The computer vision necessary for autonomous cars, for example, would be impossible without AI-enabled systems in which data from multiple sensors and cameras can be fed instantly and reliably through real-time, on-board analytics.

QUALCOMM TECH INSIGHT

“As we go to IoT, everything has connectivity, whether it’s your watch or your phone or your car. Moving the car out of its isolation into the network is going to improve the experience and safety of car travel across the board.” - Patrick Little, senior vice president and general manager, automotive, Qualcomm
More From Qualcomm

In the case of autonomous cars, that connectivity has to be lightning-fast, which requires not only more silicon devices with ever-denser circuitry but also interoperability of multiple wireless systems and seamless migration from older to newer networks. Far more demanding than even airline autopilots, autonomous cars could be endangered by a delay in connectivity of even a tenth of a second. Whether at top highway speed or in complex, city driving, safety depends on the ultra-low-latency afforded by new Wi-Fi and cellular technologies—and the even more robust 5G networks that are said to be coming by the end of the decade.

By then, experts estimate there will be at least 25 billion sensors in the much-heralded Internet of Things (IoT), and that number is expected to increase exponentially in the next five years. In the case of autonomous cars, those sensors will have to be redundant and failure-tolerant, as will the communications and analytics that draw their data together into instant on-road decisions. By the AI theorem known as “no free lunch,” meaning that individual algorithms solve only individual problems, ingenious new algorithms will need to be written—for example, to distinguish between hazardous objects and benign ones. Otherwise the brain of a driverless car might be confounded, as one prototype’s was, by the sight of a woman chasing ducks in the middle of the road, doing figure eights in an electric wheelchair.

In a world where everything will be streaming data to and from everything else, the 21st-century car will be both hub and spoke. Incoming data could include everything from home-appliance or HVAC alerts from home to minute-by-minute stock ticks from your brokerage account. Outgoing data will give car dealers, fleet owners, and manufacturers insights into their customers’ hopes and dreams.

“We are building analytic systems to really understand the data flows,” says Ken Washington, Ford’s vice president of research. “That is going to give us insight into who our customers really are—everything from where they drive to what options they use most often, what music they listen to, the temperature they set in the car.”

Swapping data with dealers will also save time and money at the shop. Many of today’s smart cars already provide basic metrics—battery charge level, tire pressure, some performance diagnostics—but as the data loop gets fuller and faster, more maintenance will become preventive rather than corrective, and some will even become wholly digital, invisible, and hands-free, just as today’s smartphones automatically update buggy apps.

The Brains
Behind the Wheel

What’s driving the autonomous-car
future is a convergence of ultra-fast
connectivity, processing power,
and machine learning

What technology drives the connected, smart car?

After more than a century of gradual improvements, the global automotive industry is accelerating toward a tipping point. Driven by consumer needs for greater safety and convenience, rapid technological breakthroughs have advanced the car of the future from sci-fi figment to imminent reality, attracting new entrants as the industry shifts its century-long focus away from the body of the car to its brain. Qualcomm has been and remains at the forefront of this innovation, working at Internet speed to bring its vision of the always-connected, electric-powered, and increasingly smart vehicle to life, in a revolution that will forever change our longstanding relationship with the automobile.

Qualcomm has been improving the connectivity, convenience, and safety of cars for more than a decade. The company developed the 3G/4G-, Bluetooth- and Wi-Fi-enabled processors that allow us to connect our smartphones to our cars so that we can stream music and make phone calls.

continue reading

It designed technologies that can detect and react to elements of the roadway environment as well as the sensor-fusion technology that lets cars handle information from multiple sources simultaneously. Building on those developments as well as its leadership in cellular and Wi-Fi technologies, the company developed vehicle-to-vehicle and vehicle-to-infrastructure communications to allow cars to communicate with everything around them, a key enabler of truly autonomous cars.

“Moving the car out of its isolation into being part of a network is going to improve the experience and safety across the board,” says Patrick Little, senior vice president and general manager for automotive at Qualcomm Technologies. “As part of a network, you're able to take advantage of all this other information.”

Qualcomm’s advanced processors support the lightning-fast and reliable real-time communications required for Internet access and connected navigation. Its Snapdragon™ 820A processor is the first to include a built-in 4G LTE modem, enabling download speeds up to 600 mbps. The Snapdragon 820A also enables high-resolution 3D mapping for navigation, machine intelligence built on neural networks for advanced driver assistance systems (ADAS), and the ability to livestream HD/4K video or games to multiple displays.

The company also developed a Connected Car Reference Platform to bring all key vehicle connectivity technologies together and to ensure a migration path to the latest – such as 5G, the next-gen mobile standard designed to deliver a unified, more capable connectivity fabric for everything, including the always-connected, autonomous vehicle of the future.

The platform supports multiple electronic-control units within the vehicle and provides security for data collected through on-board machine-learning technology. It is also designed to be future-proof, able to adapt to all the innovation that might emerge over a car’s useful life.

Beyond being more connected and safer, smarter cars will be increasingly electric, reducing carbon emissions. Qualcomm’s wireless electric vehicle charging (WEVC) technology allows electric cars to refill their batteries without being plugged in—through mats embedded in parking places and eventually under roadways. That technology could ultimately eliminate the biggest barriers to the hybrid- and electric-car market: their short range and the related inconveniences of limited battery life. Wireless charging will be key to the autonomous revolution. As Qualcomm president Derek Aberle put it in his speech to the latest Mobile World Congress: “How can you have an autonomous car that you actually have to plug in? That's pretty counterintuitive.”

All of these innovations—a quantum leap in vehicle connectivity and compute technologies—will require comprehensive integration. They could also lead to some fairly dramatic changes to highways and roads. Qualcomm’s Little, however, suggests that the company’s approach to innovation has always begun with ends, not means. “The whole process at Qualcomm, not only for automotive, but for everything we design, is all right-to-left thinking,” he says. “Instead of ‘What technology do we have that we can make important to the world,’ we think, ‘What is it we want to make that would make the world better? How can we push the future forward faster?”

Learn more about Qualcomm.

The Brains Behind the Wheel

What’s driving the autonomous-car
future is a convergence of ultra-fast
connectivity, processing power,
and machine learning

What technology drives the connected, smart car?

After more than a century of gradual improvements, the global automotive industry is accelerating toward a tipping point. Driven by consumer needs for greater safety and convenience, rapid technological breakthroughs have advanced the car of the future from sci-fi figment to imminent reality, attracting new entrants as the industry shifts its century-long focus away from the body of the car to its brain. Qualcomm has been and remains at the forefront of this innovation, working at Internet speed to bring its vision of the always-connected, electric-powered, and increasingly smart vehicle to life, in a revolution that will forever change our longstanding relationship with the automobile.

Qualcomm has been improving the connectivity, convenience, and safety of cars for more than a decade. The company developed the 3G/4G-, Bluetooth- and Wi-Fi-enabled processors that allow us to connect our smartphones to our cars so that we can stream music and make phone calls.

“Moving the car out of its isolation into being part of a network is going to improve the experience and safety across the board,” says Patrick Little, senior vice president and general manager for automotive at Qualcomm Technologies. “As part of a network, you're able to take advantage of all this other information.”

Qualcomm’s advanced processors support the lightning-fast and reliable real-time communications required for Internet access and connected navigation. Its Snapdragon™ 820A processor is the first to include a built-in 4G LTE modem, enabling download speeds up to 600 mbps. The Snapdragon 820A also enables high-resolution 3D mapping for navigation, machine intelligence built on neural networks for advanced driver assistance systems (ADAS), and the ability to livestream HD/4K video or games to multiple displays.

The company also developed a Connected Car Reference Platform to bring all key vehicle connectivity technologies together and to ensure a migration path to the latest – such as 5G, the next-gen mobile standard designed to deliver a unified, more capable connectivity fabric for everything, including the always-connected, autonomous vehicle of the future. The platform supports multiple electronic-control units within the vehicle and provides security for data collected through on-board machine-learning technology. It is also designed to be future-proof, able to adapt to all the innovation that might emerge over a car’s useful life.

Beyond being more connected and safer, smarter cars will be increasingly electric, reducing carbon emissions. Qualcomm’s wireless electric vehicle charging (WEVC) technology allows electric cars to refill their batteries without being plugged in—through mats embedded in parking places and eventually under roadways. That technology could ultimately eliminate the biggest barriers to the hybrid- and electric-car market: their short range and the related inconveniences of limited battery life. Wireless charging will be key to the autonomous revolution. As Qualcomm president Derek Aberle put it in his speech to the latest Mobile World Congress: “How can you have an autonomous car that you actually have to plug in? That's pretty counterintuitive.”

All of these innovations—a quantum leap in vehicle connectivity and compute technologies—will require comprehensive integration. They could also lead to some fairly dramatic changes to highways and roads. Qualcomm’s Little, however, suggests that the company’s approach to innovation has always begun with ends, not means. “The whole process at Qualcomm, not only for automotive, but for everything we design, is all right-to-left thinking,” he says. “Instead of ‘What technology do we have that we can make important to the world,’ we think, ‘What is it we want to make that would make the world better? How can we push the future forward faster?”

Learn more about Qualcomm.

All that feedback should lead to a shorter product cycle for automobiles. By the design principle that form follows function, “the driving dynamics become not as important as the riding dynamics,” says Lawrence Burns, former vice president of R&D and planning at General Motors, now advisor to several other developers of smart- and self-driving cars. Liberated from driving as the prime directive, he predicts, “the auto industry will go back to its fashion and design roots,” meaning a much wider range of models and competition among carmakers for advanced features, creature comforts, infotainment, and, most basic, seamless connection to home and office.

Fully autonomous cars are expected eventually to reduce the cost of car travel drastically, from 70 cents a mile today to 25 cents or less, and to save U.S. commuters 100 million hours in driving time every day.

That is not expected to translate into more car buyers, however. Major carmakers already envision a day when fleet owners will deploy autonomous cars as a service. That will be a special gift to people whose inability to drive has kept them confined and dependent, such as the blind, the disabled, and people past driving age. For many others too, individual car ownership could become an expense—and responsibility—that is easy to live without.

vision

Even for current car shoppers, digital versatility is fast replacing the allure of muscular engines and rich interiors. Unthinkable not long ago, almost half of all American teenagers said in a recent survey that they would rather have an internet connection than a car. Many of their parents would probably prefer that too.

Advances in wireless technology also promise to support the market for cars powered by electricity, unplugging them from the volatility of the petroleum-producing world and hooking them up with cleaner sources of energy, some of them renewable. If charging them were not such a clunky and cumbersome process, they could already have greatly reduced U.S. carbon emissions—and cars that have to be plugged in could hardly be called fully autonomous.

Those problems are now being solved by the development of wireless charging pads for garages and parking lots and charging systems built into roadways that will wirelessly refill batteries as the cars drive by, ending not only the electric car’s recharging issues but also its limited range.

Since vehicles don’t drink, fall asleep, or get distracted, the autonomous car’s greatest virtue will be safety. Experts predict that most fatal accidents due to driver error will be eliminated, saving the lives of tens of thousands of Americans—and as many as a million people around the world—every year.

We are still years if not decades away from the completely driverless world, but fully autonomous cars could be on the road within a decade. To remove the ambiguity from the various meanings of “autonomous”, SAE International, formerly the Society of Automobile Engineers, developed a scale from 0 to 5, 0 being no autonomy, 1 being elemental driver assistance, and 2 being “partial automation” with cruise control and assistance with lane changing, which is roughly the state of the art today. The third stage is automation monitored by the human driver, but shared responsibility makes that problematic. The fifth is full autonomy, with no feedback required or expected from an occupant of the car, even in case of emergency, but that stage faces enormous hurdles, from the phase-out of driven cars and major changes required in the roadway infrastructure to validating the code and safety requirements for a world where driven and driverless traffic are mixed. Estimates of the time it would take for test-driving to deliver statistical proof of safety in that circumstance range into centuries. New algorithms and software will need to be devised that can deliver an equally reliable assurance of safety.

QUALCOMM TECH INSIGHT

“The connected part of the car that I'm most excited about is that all these modems are going to talk to each other and know where everything is. The car will be able to take quicker, better, more decisive action than we ever could.” - Patrick Little, senior vice president and general manager, automotive, Qualcomm
More From Qualcomm

The fourth stage of automation is the goal for the next decade. At that point, autonomous cars will do everything they will do in stage five but only in designated places free from driven traffic (as autonomous buses and trains now serve airports and resorts) and on roadways where special lanes are reserved for them. As Steven Shadower, who helped start the advanced transportation technology program at UC/Berkeley, put it in a recent edition of Scientific American: “When you limit the situation in which automated vehicle systems must operate, you greatly increase their feasibility.”

The car business is only one of many industries that have been radically disrupted by digital innovation, which has unseated more than half the companies in the Fortune 500 since 2000. What has been called the “third industrial revolution”—descendant of the Industrial Revolution of the 19th century and mass production in the 20th—has advanced far more quickly than the first two, and its acceleration leaves no doubt of its continuing power to shape the global economy. More companies, industries, and business paradigms will undoubtedly be displaced before it’s finished, and its effect on the market for jobs, especially low-skill/low-pay jobs, could be profound.

Sebastian Thrun is working on that now too. He went on from winning the 2005 Grand Challenge to lead Google’s self-driving-car project. Then, in 2012, he founded an online-education company called Udacity, whose MOOCs (massive open online courses) lead to technical certifications and “nanodegrees” useful for entry-level job seekers and career advancement. Two years ago, in partnership with Georgia Tech, Udacity added an online master’s degree in computer science, which costs its students only $7,000.

“I believe we live in an age where most interesting inventions have not been made, where there are enormous opportunities to move society forward,” he told Foreign Affairs. “I'm excited to live right now. But I would rather live 20 years from now or 50 years from now than live today. It's going to be better and better.”

That’s another challenge for inventors: the inevitability of being overtaken by the always more exciting future. In the words of John Perry Barlow, a founder of the Electronic Frontier Foundation and former lyricist for the Grateful Dead: “We are all immigrants in the land of our children.”

"The Space Ahead" by Michael Murphy, 2016

READ THE SPACE AHEAD

THE SPACE AHEAD

Seeing Into
The Future With The
Technology Of Art

Michael Murphy is a seminal figure in the Perceptual Art movement, and his three-dimensional portrayal of the new world of connected cars is a prime example of it, playing with boundaries in the physical world and with the viewer’s frame of reference. “The goal of this artwork is to depict the way that the autonomous driving vehicle is communicating with the world around it,” he says. “When you look at [the work] straight on, it creates an illusion of just a seamless view of the city. And then when you walk around it, it expands, and inside all these different layers are nested graphics and icons that represent all that the car is connected to.”

After earning his MFA in the Art and Technology program at The Art Institute of Chicago, he taught at the university level for twelve years, but his fascination with technology began in childhood. ”When I was a kid, I used to fantasize about telepathy, wishing you could just close your eyes and someone could hear your thoughts. When they invented cellular phones, it felt like we had sort of figured out telepathy.”

Murphy’s artwork reveals itself in layers to portray the dazzling quantity and complexity of communications between the smart car and the world around it.

He has brought that belief in the power of technological advances to all his work, including his vision of autonomous cars. “There's some pessimism about being overly connected,” he says, “but we're a global community. I'm very optimistic about the impact [connectivity] is going to have on the human race.”

“The average driver can’t take in all the information that’s being thrown at them,” Murphy says, “Having the car as a communications and interpretive tool is an incredibly positive development.”