Mark Wheeler, Co-Founder & CTO
Autonomous driving promises to be one of the most profound technology disruptors of the 21st century. From saving lives to saving time, the advantages to society will be enormous. But the road to autonomy is not straightforward. The systems required to build and deploy autonomous vehicles (AVs) are complex, ranging from perception and localization to planning and control.
When you catch a glimpse of a self-driving car on the street, it’s easy to notice the distinct hardware on the roof, such as the hockey-puck LiDAR sensors. Less visible is the powerful mapping and localization software required for the vehicle to safely navigate and follow the rules of the road.
This is where DeepMap comes in. DeepMap was founded in 2016 by James Wu and Mark Wheeler.
Both have extensive backgrounds in digital mapping and vehicle autonomy. Previously, Wu was on the Google Earth and Maps teams and did stints at Apple and Baidu. Wheeler had been a tech lead manager at Google, where he led Google Geo enterprise products, including Google Maps Engine (GME), as well as chief software engineer at Leica Geosystems.
Today, Wheeler is DeepMap’s CTO and leader of its engineering team, while Wu serves as CEO. DeepMap’s focus is on building the technology necessary for self-driving vehicles to navigate in a complex and unpredictable environment. The company addresses three important, interrelated elements: precise high-definition mapping, ultra-accurate real-time localization, and the serving infrastructure to support massive global scaling.
“When we started the company we took a hard look at what is needed to make self-driving a real product and what that explicitly meant for maps,” says Wheeler. “We realized it required a fresh perspective on how maps are created and maintained. Rather than building on existing HD maps, our technology is designed from the ground up for a world where there are millions of autonomous vehicles continually creating and updating the maps they use.”
Wheeler explains that in AV development, mapping and localization presents one of the most difficult challenges. Maps for AVs are read by machines, not humans. A machine requires up-to-date maps, as AVs rely on real-time localization and constant updates about changes in the environment (such as road conditions, accidents, construction, and more).
Rather than collecting data with expensive survey vehicles, DeepMap has a novel approach: it leverages existing sensors in AVs to collect information and build maps, saving time and money. DeepMap’s service can be integrated into an AV’s existing planning, perception, and control systems.
“Mapping is an aspect of the self-driving stack that doesn’t always get the attention it deserves, but it is absolutely mission-critical for safety,” says Wheeler. “The AV does not need a map in the traditional sense. It needs an efficient and reliable answer to questions such as ‘where am I?’; ‘what’s around me?’; ‘what can I legally and safely do?’; and ‘how do I get from A to B?’”
DeepMap works with customers around the world. The company has a strong technical advisory board of experts who served at companies such as Apple, Google, and Uber. DeepMap’s investors include Accel, Andreessen Horowitz, and GSR Ventures.