Nvidia Creates New VR Simulation System for Their Self-Driving Vehicles

The new program, the Drive Constellation, can use sensors and make it look like the information is real. The company aims to make autonomous vehicles safe.

Nvidia has announced a new technology which will be used in autonomous driving to help to solve driving problems. The tech, also known as Drive Constellation, will be used to better the delivery of self-driving cars and keep them safe.

The senior director of the company in the Automotive Department, Danny Shapiro, said that about four to five were reported dead in two minutes for every car-related issue. Using these numbers, at least 3,000 people die globally every day because of vehicle-related incidents. He noted that the company was determined to resolve this big problem and keep the hardware and software up to par with the challenge.

Drive Constellation

The technology is a platform which can compute on two unique servers. The first server would be used to work the Nvidia Drive Sim software. This software can be used for simulation of the car’s sensors including the cameras, the lidar, and the radar. The second server will be used for the Nvidia Drive Pegasus. This is an AI automobile computer that is able to calculate and compute data from the sensors. It works with the simulated data as-if it was being fed from the sensors of that of a real car.

The Drive Sim software is able to create various photo-real information streams. With these, it can create some identical environmental conditions such as a storm, snow, high glare and many other changes. Shapiro said that it was almost like creating a hardware and software combination that would in normal cases be in the vehicle. However, in this case, it is now at the data center.

The Driver Pegasus receives the output from the simulator. It computes the information given and is able to sense the environment. In the end, it sends back the data to the simulator. These commands are sent back to the simulator at a rate of 30 per second. This allows rigorous testing without any humans put at any significant risk.

The CEO and founder of Nvidia, Jensen Huang, showed how the Drive Constellation program would work. Huang was doing his keynote, and also added that the program would be available by the end of the third quarter of this year.

Shapiro noted that the company’s stock has risen over the past few months. He said that their clientele now included anyone who was involved in making cars, trucks and even delivery vehicles. Other automobile makers who have been affiliated with Nvidia include Toyota, Volkswagen, and Volvo.

Pause on autonomous driving tests

The company also responded to the recent death of a pedestrian in the US. Elaine Herzberg was hit by an autonomous car from Uber. The accident happened in Tempe, Arizona. As a result, Uber confirmed they were suspending their self-driving tests.

A Nvidia spokesperson also told reporters that the company was also stopping their self-driving tests. The spokesperson noted that the technology used in autonomous driving was difficult, and the accident was bad. The company is trying to perfect the technology, the spokesperson added.

Huang said that they wanted to make self-driving vehicles the safest option there was.


Please enter your comment!
Please enter your name here

This site uses Akismet to reduce spam. Learn how your comment data is processed.