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

0
1685

It’s time for autonomous car builders to blaze new trails. 

NVIDIA has agreed to purchase DeepMap, a startup devoted to constructing high-definition maps for autonomous automobiles to navigate the world safely. 

“The acquisition is an endorsement of DeepMap’s unique vision, technology and people,” mentioned Ali Kani, vice president and basic supervisor of Automotive at NVIDIA. “DeepMap is expected to extend our mapping products, help us scale worldwide map operations and expand our full self-driving expertise.” 

“NVIDIA is an amazing, world-changing company that shares our vision to accelerate safe autonomy,” mentioned James Wu, co-founder and CEO of DeepMap. “Joining forces with NVIDIA will allow our technology to scale more quickly and benefit more people sooner. We look forward to continuing our journey as part of the NVIDIA team.” 

Maps which can be correct to inside a number of meters are adequate when offering turn-by-turn instructions for people. AVs, nonetheless, require a lot better precision. They should function with centimeter-level precision for correct localization, the capacity of an AV to find itself in the world.  

Proper localization additionally requires always up to date maps. These maps should additionally mirror present highway circumstances, comparable to a piece zone or a lane closure. These maps want to effectively scale throughout AV fleets, with quick processing and minimal knowledge storage. Finally, they should be ready to operate worldwide. 

Experienced Mapmakers 

DeepMap was based 5 years in the past by Wu and Mark Wheeler, veterans of Google, Apple and Baidu, amongst different firms. The U.S.-based firm has developed a high-definition mapping answer that meets these necessities and has already been validated by the AV trade with a wide selection of potential clients round the world.  

A DeepMap map in San Jose, Calif., depicting extremely detailed options of the highway and surrounding metropolis block setting, together with a dependable semantic layer of data with key attributes comparable to navigable boundaries, lane boundaries, crosswalks, visitors indicators and visitors alerts, express and implicit yield strains, and lane connectivity.

The group, primarily positioned in the San Francisco Bay Area, has many a long time of collective expertise in mapping expertise and developed an answer that considers autonomous automobiles each map creators and map customers. Using crowdsourced knowledge from car sensors lets DeepMap construct a high-definition map that’s constantly up to date as the automotive drives. 

Ongoing Partner Support  

NVIDIA will proceed working with DeepMap’s ecosystem to meet their wants, investing in new capabilities and companies for new and current companions. 

NVIDIA DRIVE is a software-defined, end-to-end platform — from deep neural community coaching and validation in the knowledge middle to high-performance compute in the car — that permits steady enchancment and deployment through over-the-air updates.  

DeepMap’s expertise will bolster the mapping and localization capabilities out there on NVIDIA DRIVE, making certain autonomous automobiles at all times know exactly the place they’re and the place they’re going. 

“We are excited to welcome the DeepMap team to NVIDIA. They have a proven track record and are entrepreneurial, nimble and engineering-focused. DeepMap meets a deep need in the industry, and together we will develop and extend these capabilities,” mentioned Kani.  

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