By Melody Li, Product Management and Jian Wu, Engineering
At Civil Maps, we are thrilled with the response to our webinar series, Getting Started With Sensor Data and pleased to announce our first Chinese-language webinars next week and in September. They’ll be hosted by our own Melody Li, Technical Product Manager for HD Semantic Maps, and Jian Wu, Research Engineer. Our CEO, Sravan Puttagunta, will also be making an appearance. If you speak Chinese, please join us!
The first session takes place Thursday, August 17th at 9:30 AM, Beijing Time, (that is Wednesday, August 16, 6:30 PM, Pacific Time). We’ll be exploring the framework we work from at Civil Maps. It is effortless to integrate with and has many advantages that can speed up autonomous vehicle development work.
Civil Maps provides cognition to autonomous vehicles via HD Semantic Maps, localization in 6D, crowdsourced edge mapping, and Augmented Reality Maps. Our streamlined data architecture is extremely lightweight, capable of reducing gigabytes of map data into kilobytes. This structure makes crowdsourcing map data on a mass scale possible, even using inexpensive ARM processors — no hard drive in the trunk required.
We’ll also talk about our techniques we use to accomplish sensor fusion with the Civil Maps hardware abstraction layer (HAL), coded using rock-solid C++ libraries. This allows for faster development work and enables us to be sensor-agnostic, capable of working with any sensor a customer desires.
At the webinar, we’ll also showcase our Atlas DevKit and Atlas DevKit Lite (use your sensors). The Atlas DevKit is a hardware+software, plug-and-play, Civil Maps system used for map creation, usage and crowdsourcing. The Atlas DevKit Lite enables developers to use their sensor configurations with our software.
The second session will take place on Thursday, September 21st at 9:30 AM, Beijing Time (Wednesday, September 20 at 6:30 PM, Pacific Time).
In this webinar, we’ll talk about Localization as an “app” and how we precisely localize vehicles as they move through a three-dimensional world. We do this in six degrees of freedom (6DoF). Such precision allows us to project an HD Semantic Map over the car’s sensor space so its computer can efficiently prioritize its attention. This process is a far less compute-intensive task than continually analyzing a real-world scene and comparing it to a stored image as other systems demand.
Next up — we’ll talk about some of the ways Civil Maps approaches sensor fusion. We aggregate the outputs of multiple sensors — typically, LiDAR, IMU, and GPS — to derive an accurate view of the environment. Our HAL allows us to do this very quickly, and the car can use the data in real time. The HD Semantic Map is also treated as an additional sensor, helping to continually validate the entire sensor suite’s output to derive a robust ground truth.
We’ll end this second session with a discussion about how partners can integrate with the Civil Maps’ mapping engine.
If you speak Chinese, please join these special webinars. Registering for is easy, and we can’t wait to “see” you online!
Melody Li is a Technical Product Manager at Civil Maps, leading the team developing the company’s HD semantics technology. In her previous role, she was a principal at Innospring, where she conducted market analysis and technical due diligence in the field of Artificial Intelligence, Robotics, and Big Data. She also gained extensive experience with enterprise integrations and focused on Computer Vision research while serving as an Associate at the investment firm, Black Rock. Melody earned a Master’s Degree in Computer Science at Carnegie Mellon University.
Jian Wu is a Research Engineer at Civil Maps. He focuses on the company’s work in sensor fusion and localization in 6 degrees of freedom for autonomous vehicles. While attending the University of California, Berkeley, he was a member of the Autonomous Lab while completing his Master’s Degree in Mechanical Engineering. As an undergraduate, he studied Mechanical Engineering and Aerospace at the University of Michigan. Jian grew up in Xinjiang province and Shanghai.
P.S. We’re hiring! Don’t see your role there yet? Email: firstname.lastname@example.org.
Also published on Medium.