Cognata | Autonomous and ADAS Vehicles Simulation

Simulation

Our autonomous and ADAS simulation platform accelerates testing and validation time to market of automated driving systems (AV) by supporting the entire development process: training, testing, validation and certification - all the way to commercial deployment

Training

Key features

Realistic Digital Twin 3D Environments

Rich drivable 3D environments for AV simulation
Digital twins or entirely synthetic worlds for AV testing and validation
Photorealism synthetic data enabled by deep neural networks (DNN)

Synthetic Datasets

Large Scale datasets for training and testing AI machine learning (ML)
Exportable and reusable
Pixel perfect, Accurate and consistent annotation in every frame

Real-life Traffic Agents

Realistic drivers, cyclists, and pedestrians
AI with geo-specific interactive behavior profiles
Configurable & deployable at scale

Validation

Key features

Ready-to-Use AV Scenarios

Library of pre-built AV and ADAS scenarios for common automated driving functions
Support for standard ADAS and AV assessment programs
Fuzzing to quickly create new variations of scenarios for regulations and certification

Scenario Authoring & Management

Authoring, configuration, and management of scenarios
Simple web-based graphical interface
Powerful Python scripting capability

Simple Integration with Existing Components

Closed-loop simulation over standard interfaces
Automated testing with Jenkins and Bamboo
Coverage-based validation for PLM and requirement management

Analysis

Key features

Extracting Insights

Ready-to-use pass/fail criteria for validation and certification of AV and ADAS
Powerful custom rule authoring
Trend mining for large scale simulation

Visualization

Goal-oriented data representation
Multi-layer adaptive dashboards
Design decision augmentation

Deployment

Measured progress toward road readiness
AV and ADAS virtual certification of regulatory requirements
Continuous support toward geo-expansion

AV Simulation: How it works

Static Layer - Digital Twin

Automatic Virtual Worlds for AV Testing

Procedurally-built 3D environments of roads, buildings, and infrastructure accurate down to the last lane marking, surface material, and traffic light.

Dynamic Layer - Vehicles and Pedestrians

Intelligent Traffic Agents for Path Planning and Policy Training

AI drivers, pedestrians, and cyclists that accurately replicate behavior customary to the region being simulated.

Sensing Layer

Accurate Simulation of Sensors

Popular camera, LiDAR, and radar photorealistic sensor models, built with deep neural networks (DNN), and a toolchain for quickly on-boarding new sensors.

Cloud Layer

Elastic Scale in the Cloud

Cost-effective infinite scale to cover thousands of scenarios and deliver results faster.

Subscribe to our newsletter