Audi Taps Israeli Tech Firm Cognata for AV Simulation Tests

Autonomous Intelligent Driving (AID), a subsidiary of German auto giant Audi, announced it is partnering with Cognata, an Israeli tech firm specializing in virtual tests for autonomous vehicle software.

Cognata’s platform leverages artificial intelligence, deep learning and computer vision within a realistic simulation environment to judge and validate AVs prior to physical roadway tests.

The company uses patented computer vision and deep learning algorithms to automatically generate a whole city-simulator including buildings, roads, lane marks, traffic signs and even trees and bushes.

The Cognata software can recreate cities from around the world, allowing an expanded range of testing scenarios, including AI-based traffic models simulating real-world traffic conditions.

The simulation engine reproduces sensor input by emulating the specific sensors’ interactions with real-world materials, while local weather conditions and lighting are added to stress test the system.

“At Audi, we are convinced that simulation is a key tool to increase our development speed and a necessary one for the validation of our product and for proving it is safe. After exploring various solutions, we decided that partnering with Cognata is the fastest way to reach these goals,” Alex Haag, AID’s chief technology officer, wrote in a statement.

The deep learning simulation engine leverages a reality-grade city mesh combined with a deep neural network and AI capabilities.

“Thorough simulation is a critical and integral part of safely preparing autonomous vehicles for the road, and we look forward to a long-term collaboration with Audi,” Cognata CEO Danny Atsmon wrote in a statement. “The combination of the groundbreaking work done by Audi and the end-to-end simulation offering from Cognata will safely accelerate commercial deployment of Audi-enabled autonomous vehicles.”

In January, AID licensed Mapillary’s Vistas Dataset to develop technology for self-driving cars across the whole Volkswagen Group.

The Mapillary Vistas Dataset includes 25,000 images that have been manually annotated into 100 different object classes from many different locations as well as different scenarios, such as time of day, weather and season.

Audi isn’t the only automaker looking into in AV simulation platforms. This week, Toyota’s autonomous-vehicle division in Silicon Valley invested $100,000 in CARLA, an open source simulation project, to take advantage of the collaborative development potential of an open-source community.

The Toyota Research Institute (TRI) also includes tools for simulating different weather and lighting conditions, plus virtual sensors such as cameras and GPS devices that can be applied to virtual vehicles, and the code and resources are available free of charge on GitHub, making them accessible to developers around the world.

Chipmaker Nvidia and Microsoft are also investing in software that can help with virtual vehicle testing.

Nvidia’s platform simulates highways to hell, replete with nightmarish scenarios of weather and bad drivers, in software and uses them to train and test autonomous driving platforms.

Meanwhile, Microsoft’s Road Runner simulator helps train AV algorithms using robotics, deep learning and hardware engineering to cloud computing and development operations to create photo-realistic driving environments.

Nathan Eddy is a filmmaker and freelance journalist based in Berlin. Follow him on Twitter @dropdeaded209_LR.


Leave a comment

Your email address will not be published. Required fields are marked *