Nvidia Proposes Self-Driving License for Autonomous Vehicles

author: Vineeth Joel Patel   

Getting vehicles to drive themselves is an expensive and tedious process, which is why the majority of traditional automakers have partnered with chipmakers like Nvidia to make the brains and the eyes of the system. During the company's 2019 GMU Technology Conference, Nvidia announced that it has developed a new version of its Nvidia Drive platform – now version 9.0. And, based on the announcement, the platform is so advanced that the company thought it was best to work with German standards body TUV SUD to introduce a driver's license for autonomous cars. 

How Advanced Is The New Version?

Drive AP2X Software 9.0, as it's called, will come out next quarter and has even more self-driving capabilities. According to Nvidia, the new software features more deep neural networks, additional sensor integration options, and facial recognition capabilities. MapNet is also included with the software and adds a DNN that's capable of identifying landmarks and lanes. 

Companies testing autonomous vehicles have found winters to be tough for self-driving cars, as the slush, bright reflection of the sun, and debris that covers sensors throws the systems off its game. That shouldn't be a problem with the Drive AP2X platform, which has ClearSightNet that allows the driverless car to "make up for any sensor obstructions." 

Another new feature includes the platform's ability to generate its own map without having to utilize an existing form of guidance to get around. Companies usually scout out specific sections to gather data to give autonomous cars high-tech maps to go off of, but Drive 9.0 can generate its own by using all of its hardware components to create what the Nvidia calls a Local HD map. 

It's All About Safety

The most notable thing about Drive AP2X, though, is Safety Force Field. The moniker refers to an open-source collision avoidance algorithm that predicts where nearby obstructions will pop up. Alongside making predictions, the algorithm actively acts to shield the self-driving car from getting into a collision by using the latest safety features, including automatic emergency braking and intelligent steering.  

Interestingly, CNET reports that Nvidia is proposing that autonomous vehicles, in a similar fashion to human teenagers, should pass a series of tests to earn a "driver's license." The tests would be completed using Nvidia's latest Drive Constellation software and have the company's Safety Force Field algorithm as the foundation. 

Seeing as how Nvidia Drive 9.0 will offered to automakers and other tech companies in the near future, and that Safety Force Field is open-source, having automakers put their vehicles through a series of tests that's similar to a "driver's license" is a sound idea that could sway the public into trusting the vehicles more. 

Vineeth Joel Patel

Joel Patel has been covering all aspects of the automotive industry for four years as an editor and freelance writer for various websites. When it comes to cars, he enjoys covering the merger between technology and cars. In his spare time, Joel likes to watch baseball, work on his car, and try new foods