Civil Maps, a developer of cognition systems for autonomous vehicles, has already secured new OEM partnerships, just one month after publicly announcing its AtlasDevKit platform for localization, Mobility Buzz has learned.
Civil Maps’ cognition system — an AI system that allows a car to understand or see its surroundings — enables autonomous vehicles to crowdsource high-definition, machine-readable dynamic maps for safer driving. The company released its Atlas DevKit on March 7, which is a hardware and software offering that enables the low-cost, real-time creation and conversion of sensor data into 3D semantic map data.
Traditionally, low-level localization and data collection can incur hardware costs alone in the $100,000 to $500,000 range, while requiring significant computing power, according to the company. Civil Maps offers its Atlas DevKit platform in packages where the hardware bill-of-material (BOM) starts at that start at $20,000 as part of an R&D contract with the company. A BOM is a list of the raw materials, sub-assemblies, intermediate assemblies, sub-components, parts and the quantities of each needed to manufacture an end product.
This R&D contract could include car manufacturers, trucking companies, or any company with a large fleet of cars, Fabien Chraim, vice president of research at Civil Maps, told Mobility Buzz.
Chraim would not specify which, or how many, newly partnered OEMs are using Civil Maps’ DevKit. San Francisco-based Civil Maps currently has partnerships with “several” OEMs in Europe, the United States, China, and Japan.
“We have been having conversations [with potential customers] … it’s an ongoing relationship,” Chraim said. “We aren’t looking for one-time partnerships, but working together [continuously].”
The idea behind R&D partnerships through the DevKit is that the customers can “play” with the technology and give Civil Maps insights into how to improve it, he added.
The goal is to demonstrate the company’s capabilities, Chraim said. The DevKit allows Civil Maps to show its ability to reduce the cognitive load that autonomous vehicle technologies currently use, by analyzing real-time data on small processors, instead of large servers and other bulky equipment. Several OEMs have approached Civil Maps because of this, he added.
“If you look at deployment today [of autonomous vehicle fleets] … the reality is that these cars are using a lot of [expensive] sensors … [and] because you are producing so much data in real time, this data needs to be analyzed and context extracted from it,” Chraim said. These cars will have large servers and cooling systems in the back of the car in order to store the data for later, he added.
Instead, Civil Maps and the Atlas DevKit, are able to process the data on the spot, and compress the gigabytes of raw sensor input down to a few kilobytes of (map) information per kilometer. “The result is basically that you can fit the map of a city in an email attachment,” Chraim said.
Civil Maps, founded in 2013, focuses on mapping the static infrastructure of roads, including road edges, signs, and signals, Chraim said.
The hardware aspect of the DevKit includes third-party sensors such as lidar, cameras, inertial measurement systems, and communication devices. But despite the hardware components added to the Atlas DevKit, Civil Maps is not looking to operate its own fleet at any point, Chraim said. Because of this, there is also an alternative version of the platform– the Atlas Lite DevKit — which integrates with an autonomous vehicle’s existing sensors.
Those interested in learning more about the evolution of the transportation industry should join us at the second annual Auto Finance Innovation 2017 conference, May 17-18 at the Hilton Bayfront in San Diego. Visit www.autofinanceinnovation.com and to learn more.2 - Readers Like This Post