Follow
Subscribe

Bosch Invests in Autonomous Driving Map Startup DeepMap

Home > News > Content

【Summary】Robert Bosch Venture Capital GmbH (RBVC), the venture capital arm of global automotive parts supplier Bosch Group, has completed an investment in mapping startup DeepMap Inc, a start-up based in Palo Alto, California that is building high definition maps specifically for self-driving vehicles.

Eric Walz    Sep 20, 2018 2:24 PM PT
Bosch Invests in Autonomous Driving Map Startup DeepMap

Robert Bosch Venture Capital GmbH (RBVC), the venture capital arm of global automotive parts supplier Bosch Group, has completed an investment in mapping startup DeepMap Inc, a start-up based in Palo Alto, California that is building high definition maps specifically for self-driving vehicles.

DeepMap is focused on solving the mapping and localization challenge for autonomous vehicles. The investment amount was not disclosed.

"Maps explicitly designed to be read by machines are a critical enabling technology for safe autonomy. DeepMap fills a vacuum in the market. The company's approach to mapping, which leverages embedded software on the vehicle, is very compelling and relevant for highly automated as well as autonomous driving, within Bosch and the whole automotive industry," says RBVC Managing Director Dr. Ingo Ramesohl. "DeepMap builds upon the growing RBVC portfolio of technologies addressing the autonomous vehicle vertical, next to start-ups like AImotive and Tetravue."

DeepMap is taking a unique approach within the mapping industry and allows different companies to create custom high definition (HD) maps, which contain different characteristics and data. DeepMap also provides software embedded on the vehicle that addresses localization, data updates, route planning and data collection.

Deepmap's high definition (HD) maps are complete and accurate reconstructions of a 3D environment. The HD maps provide a critical set of "pre-information" to autonomous vehicles. They contain layers of semantic information such as lane connectivity, drivable areas, pedestrian crosswalks, road signs, traffic signals and other information that is specific to each end customer's solution.

HD maps have extremely high precision— down to just a few centimeters. The reason is that autonomous vehicles need very precise instructions on how to maneuver in the real world. The more precise the HD map is, the simpler the self-driving software can be, while obtaining higher performance.

"Self-driving car development is both a technology challenge and a product challenge. It needs to be safe, affordable, and scalable. In the self-driving era, a map is no longer 'just a map'. It's an integral part of the car's brain," said James Wu, CEO and Co-Founder of DeepMap.

DeepMap's technology fuses images from digital cameras with data collected using LiDAR to build detailed 3D maps. Using LiDAR and camera technology, DeepMap can identify street signs, mailboxes, even small details such as the height of a curb.

localintersection2_img_h720.png

DeepMap builds maps using using 3D lidar data

The 3-D maps complement a self-driving vehicle's sensor suite, by giving the cars a detailed awareness of the environment outside the car's view. DeepMap intends to provide a high-definition mapping and localization service to support millions of cars.

Deepmap is using a unique approach to keeping maps updated. Typically, maps are made by manually collecting data by driving a mapping vehicle on public streets until sufficient data gathered to built the map. Once the map is completed it can be uploaded to a driverless vehicle to use.

However, the challenge in creating maps this way is that once the map is completed it is no longer ‘live'. For example, suppose a road is closed for construction, a self-driving vehicle needs this updated map information, or it might get stuck trying to navigate through a construction zone that the vehicle does not know exists.

DeepMap is addressing this specific challenge, providing ‘live', continuously updated HD maps. The company is providing a full software layer in the cloud and embedded on the vehicle.

In addition, DeepMap is pursuing a new data ownership model where the end customer retains ownership of the underlying data. Instead of owning and operating expensive and expensive fleet of mapping vehicles, data can be collected from the customer's autonomous vehicles that are already deployed in the field to keep the maps updated.

Deepmap's team has extensive experience at launching maps at massive scale. Members of the company's team have previous experience working for Google Maps, Google Earth, and the Baidu HD Mapping initiative.

RBVC joins previous investors Andreessen Horowitz, Accel Partners, and GSR Ventures as major investors in DeepMap.

Prev                  Next
Writer's other posts
Comments:
    Related Content