NVIDIA to Acquire DeepMap, Enhancing Mapping Solutions for the AV Industry


It is time for autonomous motor vehicle builders to blaze new trails. 

NVIDIA has agreed to get DeepMap, a startup dedicated to building higher-definition maps for autonomous vehicles to navigate the globe safely and securely. 

“The acquisition is an endorsement of DeepMap’s unique eyesight, know-how and men and women,” reported Ali Kani, vice president and general manager of Automotive at NVIDIA. “DeepMap is envisioned to extend our mapping items, help us scale all over the world map operations and expand our full self-driving knowledge.” 

“NVIDIA is an awesome, globe-shifting firm that shares our vision to speed up safe autonomy,” stated James Wu, co-founder and CEO of DeepMap. “Joining forces with NVIDIA will permit our engineering to scale additional quickly and benefit a lot more folks quicker. We glimpse forward to continuing our journey as aspect of the NVIDIA team.” 

Maps that are precise to in just a several meters are great enough when supplying transform-by-change instructions for individuals. AVs, nevertheless, need a great deal higher precision. They should operate with centimeter-level precision for correct localization, the capacity of an AV to identify itself in the globe.  

Proper localization also needs constantly current maps. These maps should also reflect current highway situations, these types of as a work zone or a lane closure. These maps will need to successfully scale throughout AV fleets, with rapidly processing and negligible information storage. Eventually, they must be able to perform worldwide. 

Expert Mapmakers 

DeepMap was founded 5 several years ago by James Wu and Mark Wheeler, veterans of Google, Apple and Baidu, among the other organizations. The U.S.-based firm has designed a large-definition mapping answer that fulfills these specifications and has currently been validated by the AV industry with a huge array of prospective shoppers around the globe.  

A DeepMap map in San Jose, Calif., depicting extremely-in-depth characteristics of the highway and bordering town block environment, such as a reputable semantic layer of info with essential characteristics this kind of as navigable boundaries, lane boundaries, crosswalks, visitors indications and site visitors signals, express and implicit yield traces, and lane connectivity.

The workforce, mostly located in the San Francisco Bay Area has a lot of many years of collective experience in mapping technologies, and developed a remedy that considers autonomous cars both equally map creators and map buyers. Utilizing crowdsourced facts from auto sensors lets DeepMap build a significant-definition map that’s continuously updated as the auto drives. 

Ongoing Spouse Support  

NVIDIA will continue working with DeepMap’s ecosystem to meet their needs, investing in new abilities and solutions for new and existing partners. 

NVIDIA Travel is a software program-defined, conclusion-to-conclusion platform—from deep neural community teaching and validation in the info centre to significant-overall performance compute in the vehicle—that enables continuous improvement and deployment via about-the-air updates.  

DeepMap’s technology will more bolster the mapping and localization capabilities accessible on NVIDIA Push, guaranteeing autonomous vehicles always know specifically the place they are, and where by they are heading. 

“We are psyched to welcome the DeepMap team to NVIDIA. They have a tested track file and are entrepreneurial, nimble, and engineering-concentrated. DeepMap meets a deep have to have in the marketplace, and together we will create and lengthen these capabilities,” explained Kani.  

The acquisition is expected to shut in the third calendar quarter of 2021, subject to regulatory approval and customary closing circumstances. 

Leave a comment

Your email address will not be published.