Sustainability

Subscribe
Safer Roads, Connected Car

Autonomous Cars

Safer Roads, Connected Cars

 
Autonomous cars could change the world
for better

Autonomous driving is quickly becoming a reality, and it's bringing with it far reaching social and environmental impact.

NVIDIA plays a leading role in developing technology and software that makes self-driving vehicles possible. And, by making driving more environmentally sustainable and potentially saving millions of lives, our efforts are in line with the United Nations Sustainable Development Goals of fostering innovation and sustainable cities and communities.

We do this by tapping into the innovative work of our employees worldwide to develop cutting-edge technology and by maintaining partnerships with leading companies across the automotive industry.

"Autonomous vehicles hold enormous promise to modernize our transportation system, and AI has the power to make them a reality," said Jensen Huang, NVIDIA founder and CEO, in his opening keynote address at the 2017 Consumer Electronics Show, the world's most important technology tradeshow.

 
 
Transit Change

Self-driving cars will make driving safer, reduce carbon emissions, and transform urban transit. But there's an even bigger story here that's deeply rooted in how artificial intelligence is helping build a fully connected car.

Advances in AI are creating training software to ensure that cars are smart enough to safely drive themselves. They're also enabling cars to recognize the most-subtle experiences in the environment around them, and to be highly tuned to the people traveling inside.

 
 
$1 Trillion in Social, Consumer Benefits

Self-driving cars could account for 21 million new vehicles on the road by 2035, according to research group IHS Automotive. That's thanks in part to the United States' broad investments in research and development by tech companies, automakers, and parts suppliers.

Automakers and tech companies racing to develop self-driving cars, and related services such as ride-sharing, are equally likely to work together to develop the technologies needed to advance autonomous vehicles.

The shift to autonomy could deliver $1 trillion in social and consumer benefits, in large part due to improving safety rates.

With deep learning, cars can now perceive their surrounding environment and create incredibly precise HD maps that enable them to orient themselves, understand the world around them and predict where things will be in the immediate future. Supported by HD maps in the cloud, continuously updated as the world changes around the car, allows vehicles to plan a safe path forward.

Global HD mapping services offered by Baidu, HERE, TomTom and Zenrin have all adopted NVIDIA MapWorks for map processing and live map updates in the car. And the NVIDIA DRIVE PX AI car supercomputer performs continuous mapping.

 
 
Saving Lives, Cutting Emissions

Even with gains in car safety, the rise in motor vehicle deaths made 2016 the deadliest year on the nation's roads in almost a decade, National Safety Council figures show. Almost 40,000 people were killed on U.S. roads, with 4.6 million seriously injured, in part because of increased risky behaviors such as speeding and texting while driving.

Safety issues focus on drivers, passengers, and pedestrians. If even just a few vehicles worldwide have some self-driving capability within the next decade, the technology could reduce or even eliminate driver error, saving countless lives.

Fuel use also could be cut, leading to lower carbon emissions. Adoption of autonomous vehicles by large transit fleets would improve public transportation efficiencies, tackle complex and costly infrastructure issues, and potentially transform sprawling parking lots into parkland.

The trajectory of such profound changes is outlined in a report by consulting firm McKinsey & Co., which notes the benefits of autonomy include labor-saving costs and reduced carbon dioxide emissions through optimized driving.

 
 
Transformative Power of Deep Learning

Driverless cars weren't possible until researchers adopted an AI technique known as deep learning. It relies on the powerful parallel processing ability of GPUs, access to vast troves of data, and sophisticated algorithms, called deep neural networks, to solve complex problems.

Deep learning is essential for autonomous vehicles because it's impossible to write software that anticipates every possible scenario a self-driving car might encounter. It enables the car's computer to learn, adapt, and improve.

Now, almost every major carmaker and ride-sharing firm, as well as numerous startups and tech companies, are developing cars that can drive themselves some or all of the time - developments that won't only transform how people get around, but improve auto safety.

Schoolchildren will be able to get to sports events in a timely way without a parent driving them. The elderly will be able get to medical appointments on their own. Commuters will be able to use their time productively on their way to work. And those who can't drive will find themselves empowered with mobility without ever taking a driving lesson.

Powered by supercomputers and programmed for safety, driverless cars would avoid such human errors as speeding, drunken driving, and running red lights. Moreover, computers don't get distracted by phone calls or texts, and they never get drowsy.

 
 
Autos and Beyond

Mercedes-Benz, Audi, Toyota, and Volvo are some of the carmakers working on self-driving vehicles with safety as a key goal, and autonomous features that make traveling a great experience for drivers and passengers inside vehicles.

Sweden's Volvo is piloting a fleet of self-driving SUVs using NVIDIA DRIVE PX technology in its hometown of Gothenberg. Audi first showed off its self-driving capabilities two years ago with a vehicle that drove itself several hundred miles from the San Francisco Bay Area to Las Vegas. Mercedes-Benz expects to have a vehicle on the road utilizing NVIDIA technology within the coming year.

Toyota aims to make its self-driving cars safer by using GPU-powered simulation and deep learning. These technologies will allow autonomous vehicles to experience and train on numerous real-world situations before they're on the road.

Equipping cars with the NVIDIA DRIVE PX 2, a supercomputer for vehicles that uses deep learning, sensor fusion, and surround vision to understand a car's surroundings and navigate a safe path forward, is helping automakers advance their vehicles' learning.

NVIDIA is also collaborating with Bosch, the world's leading tier-one automotive supplier supporting every car maker in the world, to build an AI car computer for the mass market.

The already powerful DRIVE PX platform will soon incorporate NVIDIA's forthcoming Xavier technology, which can process up to 30 trillion deep learning operations a second while drawing just 30 watts of power.

 
 
Industrial Applications

AI is set to revolutionize the $10 trillion transportation industry, a cornerstone of our society's infrastructure. There are 300 million trucks on the road that drive 1.2 trillion miles per year. Adding autonomy to this industry will improve driver productivity, enhance transportation efficiency, and increase safety.

Working with PACCAR, a leading global truck manufacturer that makes the Kenworth, Peterbilt, and DAF lines of trucks, is an opportunity to explore AI and automated driving systems.

One of the industry's largest vehicle suppliers, ZF, this year launched the ZF ProAI self-driving system, based on the NVIDIA DRIVE PX 2 AI car computer, for cars, trucks, and factories. Their goal is to take automated and autonomous driving systems beyond self-driving cars to industrial applications such as forklift trucks and materials-handling vehicles.

Autonomous vehicles hold enormous promise to modernize many aspects of our transportation system, and AI has the power to make them an everyday reality.