Scenario Builder for Automated Driving Toolbox

Build simulation scenarios from real-world vehicle data recorded from GPS, IMU, camera, and lidar sensors

The Scenario Builder for Automated Driving Toolbox, allows users to generate simulation scenarios for automated driving applications. It provides functions that helps to generate scenarios from both raw real-world vehicle data and processed object list data from perception modules.

With RoadRunner Scene Builder, users can visualize the generated scenes with roads, lanes, and roadside objects in RoadRunner. These scenes can then be exported to ASAM OpenDRIVE® format. On the other hand, RoadRunner Scenario tool facilitates the simulation of scenario actors, such as vehicles, along their trajectories. The resulting scenarios can be exported to ASAM OpenSCENARIO® v1.x and v2.0 formats.

The generated scenes and scenarios can be used for designing and testing automated driving applications.

Getting Started

• Overview of Scenario Generation from Recorded Data
• Documentation
• Examples

Features

Scenario generation using recorded sensor data is accomplished using below workflow:

Key event extraction

• Extract key events from recorded sensor data

Ego vehicle trajectory generation and localization

• Smooth GPS Waypoints for Ego Localization
• Ego Vehicle Localization Using GPS and IMU Fusion
• Ego Localization Using Lane Detections and HD Map

Road reconstruction

• Generate RoadRunner Scene Using Labeled Camera Images and GPS Data
• Generate RoadRunner Scene Using Labeled Camera Images and Raw Lidar Data
• Generate RoadRunner Scene from Recorded Lidar Data
• Generate High Definition Scene from Lane Detections and OpenStreetMap
• Extract Lane Information from Recorded Camera Data for Scene Generation
• Preprocess Lane Detections for Scenario Generation

Roadside object reconstruction

• Generate RoadRunner Scene with Trees and Buildings Using Recorded Lidar Data
• Generate RoadRunner Scene with Traffic Signs using Recorded Sensor Data
• Generate RoadRunner Scene Using Aerial Lidar Data
• Generate RoadRunner Scene Using Aerial Hyperspectral and Lidar Data
• Georeference Sequence of Point Clouds for Scene Generation
• Georeference Aerial Point Cloud for Scene Generation

Target actor trajectory reconstruction

• Generate RoadRunner Scenario from Recorded Sensor Data
• Generate Scenario from Actor Track List and GPS Data
• Fuse Prerecorded Lidar and Camera Data to Generate Vehicle Track List
• Extract Vehicle Track List from Recorded Camera Data
• Extract Vehicle Track List from Recorded Lidar Data
• Extract 3D Vehicle Information from Recorded Monocular Camera Data for Scenario Generation

Overview Video

Generate Scenarios from Recorded Sensor Data - YouTube

Technical Article

APTIV - Harvesting Driving Scenarios from Recorded Sensor Data

ARAI - ARAI Enables Virtual Testing of ADAS Application with Real-World Simulation Scenarios

Customer Reference

Aptiv:

Scenario Harvesting Using Automated Driving Toolbox and RoadRunner Scenario, MathWorks Automotive Conference 2023, NA

Scene Sync: Bridging Real-World Scenarios with Virtual Environments for ADAS Development, MathWorks Automotive Conference 2024, NA

IAV:

AD/ADAS Country-Based Virtual Validation Using Real-World Data, MathWorks Automotive Conference 2024, Europe

Ashok Leyland:

Ashok Leyland leverages Gen AI functionality to generate scenarios with India specific assets for ADAS perception validation, MathWorks Automotive Conference 2024, India

Higer:

Higer reconstructed scenarios for Autonomous Shuttle validation using Simulink, MathWorks Automotive Conference 2024, China (English Translated Slides)

Support

If you have download or installation problems, please contact Technical Support.

Notes

You can create multiple variations of a generated scenario to perform additional testing of automated driving functionalities. For more information, see Scenario Variant Generator for Automated Driving Toolbox.