According to Jensen Huang, founder, and CEO of Nvidia, the business has unveiled a new mapping platform that will give the autonomous car industry with ground truth mapping coverage of over 300,000 miles of highway in North America, Europe, and Asia by 2024. The Drive Map platform is designed to enable high degrees of autonomous driving. Drive Map is not just available to existing Nvidia customers, but it also complements the company’s existing AV products.
At the same event, Nvidia unveiled the latest iteration of Drive Hyperion, the company’s sensor and compute self-driving toolkit utilized by Mercedes, Volvo, JiDu, and, as of Tuesday, BYD and Lucid Motors to deliver a variety of smart driving and enhanced driving support functions. TuSimple, WeRide, Zoox, and DeepRoute.ai are some of Hyperion’s customers.
Drive Map is the result of Nvidia’s acquisition of DeepMap, a high-definition mapping firm, last year. By combining DeepMap’s precise survey mapping with anonymous mapping data crowdsourced from all cars that employ Nvidia’s Hyperion architecture, the solution achieves centimeter-level precision. Three localization layers — camera, lidar, and radar — are included in the mapping tool to give the redundancy required for autonomy.
As cars travel, all of the data collected from Nvidia customers is continually transferred to the cloud. It’s then compiled and put onto Nvidia’s Omniverse, the company’s open platform for virtual collaboration and real-time physically realistic simulation, where it’s utilized to update the map so that cars can properly locate themselves. Nvidia is able to increase its mapping footprint more quickly as a result of this.
Omniverse also use automated content creation technologies to create a comprehensive map, which is then transformed into a drivable simulation environment that can be utilized with Nvidia Drive Sim, an end-to-end autonomous vehicle simulation platform. Nvidia has announced the acquisition of DeepMap, a high-definition mapping business. Nvidia Drive, the firm’s autonomous car technology division, will benefit from the mapping IP, according to the business.
In a statement, Ali Kani, vice president and general manager of Nvidia’s automotive division, stated, “The purchase represents an affirmation of DeepMap’s unique vision, technology, and people.” “DeepMap is expected to help us build our comprehensive self-driving capabilities, help us scale our global map operations, and extend our mapping solutions.” One of the most difficult aspects of establishing complete autonomy in a passenger car is obtaining accurate localization and up-to-date mapping data that represents current road conditions. Nvidia’s autonomous stack should have more precision after using DeepMap’s technology, providing the car more ability to find itself on the road.