Cognata ADAS Simulation

Cognata's Supervised Generative AI Simulation Platform integrates real-world data with advanced synthetic data generation, utilizing state-of-the-art 3D rendering. This unified platform accelerates the development and validation of autonomous systems, including AVs, ADAS, drones, and robots.

Key Solutions:

• SimCloud Simulation Platform: Supports the entire development process for AV and ADAS systems, from training and testing to validation, certification, and commercial deployment.
• DriveMatrix Supervised GenAI: Transforms existing test drive videos into diverse variations under varied weather and lighting conditions using supervised Generative AI. This optimizes real data augmentation, reduces the need for additional test drives, and ensures comprehensive training. DriveMatriX can operate as a fully managed service.

Features:

• Sensors: Supports a wide range of sensors including RGB HD cameras, Fisheye cameras, Radar, LiDAR, LWIR (Thermal), and Ultrasonic cameras.
• Simple Integrations: Offers model interfaces for Simulink and ROS, tool interfaces for SUMO, automated testing with Jenkins and Bamboo, and supports OpenDrive and OpenScenario standards.
• Analytics Platform: Provides ready-to-use pass/fail criteria for validation and certification, custom rule authoring, trend mining, and virtual certification for regulatory requirements.
• Digital Twin: Features a library of pre-built scenarios, diverse terrain types, realistic vehicle physics, and a comprehensive assets catalog including vehicles, pedestrians, cyclists, and buildings.
• Vehicle Physics: Includes over 1,000 adjustable parameters for specific vehicle models and dynamics, custom 3D asset import, and a range of ego vehicle heights.
• DNN Photorealism: Generates photorealistic synthetic data using deep neural networks (DNN), offering pixel-perfect, accurate, ground truth consistent annotation and AI with geo-specific interactive behavior profiles.