Engineering Updates

A Day at the Self-Driving Races


This weekend, we’ll be heading over to Thunderhill Raceway in Willows, CA for the Self- Racing Cars event, with the goal of ultimately providing cognition and 3D mapping for autonomous competition vehicles. The two-day race was launched just last year and is gaining an enthusiastic following from local autonomous vehicle technology companies. Bringing together a “who’s who” of industry players, from self-driving car makers to sensor manufacturers, camera companies, and software development firms — all in one place, the Self Racing Cars event should also be great for networking, meeting & greeting. The actual on-track competition pits cars against cars with — no drivers and/or no physical human intervention. It is timed and scored for best autonomous laps around a 2.1-mile track. While doing a long survey drive last year, our team stopped by the race and went for a test drive.

While the races start on Saturday 4/1, we’ll bring one of our survey cars on Sunday 4/2 to create a localization reference table and 3D semantic map. The map will be accessible for racing teams at future events and can be used with low compute (ARM processor) requirements, an option that our auto OEM customers have been using with positive feedback. Self-driving cars using our software stack will gain a key advantage: localization in 6 degrees of freedom (like airplanes) and the utilization of our Augmented Reality Map to understand the environment and respond to the road conditions while driving at high speeds.

How do advanced semantic maps enhance the racing experience?

Taking a glimpse into the future — let’s examine the intersection of the autonomous vehicles industry and its intriguing implications for the world of motorsport.

Our technology is much more powerful than conventional 2D mapping applications that most of us use today. The key benefit of Civil Maps’ system is a cognition layer for the self-driving car —”enhanced information “ (including prior memory and real time data) that the vehicle’s decision engine can use to drive. Leveraging precise localization in six dimensions and our 3D semantic map, Civil Maps can give race cars an edge in maneuvering the track. Our high definition, AR maps are effectively projected into the car’s sensors to contextualize what they “see”, which gives it a powerful safety and navigation tool; Civil Maps system helps the car understand which real-world features to focus upon and how to prioritize reactions to the environment in real-time.

If you’re familiar with rally racing ( as seen in the video above), you’ll know that the driver is assisted by a navigator with “pace notes”, informing the driver with details about what’s ahead and where they should devote their attention. That conventional, location-based information system is operationally similar to what Civil Maps uniquely provides, but seems very retro in comparison.

Provided with augmented reality technology and artificial intelligence, the future of racing could look quite different if these advancements are desired and permitted to compete.

With AI assistance, cars utilizing different levels of autonomy can leverage AI and augmented reality maps to navigate and drive even faster, more daringly, and also generally safer than what is seen today. (Think Wipeout!)

While our technology is best challenged in dense urban environments, we are excited to share some of what we’ve been working on with this specialized racing community. If you’ll be there on Sunday please reach out and say hello! We’ll be in the Civil Maps car.

P.S. We’re hiring! Don’t see your role there yet? email: jobs@civilmaps.com