You have brains in your head. You have feet in your shoes. You can steer yourself in any direction you choose. You're on your own, and you know what you know. And you are the guy who'll decide where to go.
I love this quote by Dr. Seuss, but it won't be as true when driverless cars come along. When they eventually roll onto the streets en masse, they'll use their artificially intelligent brains to steer us in the direction they choose.
And it's for this reason that some of the most important technology to come out of autonomous driving research is the machine learning software and processors that will make driverless cars ubiquitous.
But behind some of Google's driverless car advancements lies another company. Chipmaker NVIDIA (NVDA 3.66%) has used its Tegra processors to help power some of Google's driverless car tech over the past few years, and the company recently announced its second generation of autonomous car software technology.
NVIDA calls its new Drive PX 2 driverless platform the "world's first in-car artificial intelligence supercomputer."
The platform can process 24 trillion deep learning operations every second.
Drive PX 2 combines the company's processors with a platform of sensors, cameras, lidar, and radar that gives autonomous vehicles the situational awareness that humans take for granted. This awareness is one of the key ingredients in making cars that can drive better than we do.
The situationally aware auto
To understand how important this is, imagine yourself driving down a neighborhood road. There are houses all around you, and suddenly a basketball rolls quickly out into the road. When you see the ball, your brain automatically tells you that a kid may be following behind that ball, and that you should probably slow down. All of that, the neighborhood, the houses, the ball, and the idea of the possible child in tow is all situational awareness.
Our experiences and years of driving put all these facts together without us even realizing it, and computers have to learn these things as well. They can be programmed, of course, but they also need deep learning to take their driving experiences and apply them to future situations and to be able to share them with other vehicles.
And that's why NVIDIA's driverless platform play is so important. The company believes that its Drive PX system will eventually have superhuman levels of situational awareness and make self-driving cars safer than if people were were behind the wheel.
The company is partnering with more than just Google to get its driverless brain power into the market. Its first-generation Drive PX was used in an Audi A7 sedan that drove 550 miles from San Francisco to Las Vegas almost entirely on its own.
But it's not just Audi that's tapping into NVIDIA's driverless tech. More than 50 carmakers and auto parts suppliers are using and testing the company's Drive PX platform, including BMW, Daimler, and Ford.
IHS Automotive predicts that by 2035 10% of light vehicles sold will be fully autonomous. Carmakers are looking to tech companies to deliver the hardware and software necessary to power the future of their vehicles, and NVIDIA's already delivering. The company's technology will likely be the brains behind our cars, steering them where to go, and deciding the safest path to our destinations.