Cognata offers a digital twin-based simulation platform that generates synthetic data and augments real-world datasets to enhance the training, testing, and validation of autonomous vehicles and advanced driver-assistance systems (ADAS). This technology addresses the challenge of insufficient and diverse data for effective development and certification of automated driving systems, accelerating their time to market.
Funding
$27.8M raised to dateRaised to date based on public sources. This may differ from the amount the company actually raised and is based only on what is publicly available on the internet.
Founders
Product
Problem
The development and validation of autonomous vehicles (AVs) and advanced driver-assistance systems (ADAS) require vast amounts of diverse data, which is often insufficient and costly to acquire through real-world driving. This data scarcity hinders effective training, testing, and certification of automated driving systems, delaying their deployment.
Solution
Cognata provides a simulation platform that leverages digital twin technology and supervised generative AI to generate synthetic data and augment real-world datasets, addressing the data challenges in AV and ADAS development. The platform supports the entire development process, from training and testing to validation and certification, accelerating the time to market for automated driving systems. Cognata's DriveMatrix solution uses existing test drive videos to create numerous variations under diverse weather and lighting conditions, seamlessly integrating into data pipelines to enhance real data efficiently. The platform's cloud architecture supports large-scale simulation, enabling the rapid creation of novel scenarios and improved AI/ML training.
Target Audience
The primary audience includes automotive manufacturers, Tier 1 suppliers, and technology companies developing autonomous vehicles, ADAS, and other automated driving systems.
Features
- Digital twin environments with pre-built scenarios, diverse terrain types, and vehicle physics behavior
- Comprehensive asset catalog including vehicles, pedestrians, cyclists, vegetation, and buildings
- Support for multiple sensors and sensor fusion, including RGB HD cameras, fisheye cameras, radar, LiDAR, LWIR thermal cameras, and ultrasonic cameras
- Simple integrations with existing components through model interfaces like Simulink and ROS, and tool interfaces like SUMO
- Automated testing with Jenkins and Bamboo, and support for standards like OpenDrive and OpenScenario
- Analytics platform with ready-to-use pass/fail criteria for validation and certification, custom rule authoring, and trend mining for large-scale simulation
- DNN photorealism enabled by deep neural networks, providing pixel-perfect, accurate, and ground truth consistent annotation in every frame
- DriveMatrix GenAI data augmentation to create diverse datasets from single drives