Virtual Traffic Scenario Generation With Domain Adaptation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Testing and validating autonomous driving systems in vehicles is complex and requires extensive real-world testing, and securing training data is difficult, especially when updates are needed, with conventional virtual environments lacking variety in scenarios.

Innovation Solution

A method and device for generating elaborate and varied traffic scenarios in a virtual driving environment using data augmentation and a scenario augmentation network with LSTM generators and discriminators, based on actual driving data, to simulate realistic driving conditions.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If real-world road testing is used for training and validating autonomous driving systems, then sufficient training data and validation coverage can be obtained, but the testing process becomes highly complex and requires prolonged time (millions of hours and miles)

Engineering Contradiction:
Improvevalidation coverageVSAvoidtesting duration
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent creates virtual copies of real-world driving scenarios by capturing actual driving data and reconstructing it in a virtual environment. This allows comprehensive validation without physical road testing, as the virtual scenarios replicate real-world conditions including rare edge cases that would be difficult to capture in limited physical testing

Inventive Principle:
Principle #26Copying

Solution Approach 2:

The system performs preliminary data collection and scenario generation before actual testing needs arise. By pre-capturing diverse driving scenarios and pre-generating virtual traffic scenarios with various conditions, the system prepares comprehensive training and validation data in advance, eliminating the need for prolonged on-road testing

Inventive Principle:
Principle #10Preliminary action

2Adaptability or versatility

If conventional virtual driving environments use only specific traffic conditions, then the environment is easier to manage, but it cannot provide various sophisticated scenarios similar to actual driving environment

Engineering Contradiction:
Improvescenario varietyVSAvoidenvironment complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent dynamically changes multiple parameters of traffic scenarios including vehicle types, pedestrian behaviors, weather conditions, road types, and traffic patterns. By systematically varying these parameters based on real driving data, the system generates diverse sophisticated scenarios while managing complexity through structured parameter control

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The virtual driving environment is designed to serve multiple functions simultaneously: training autonomous vehicles, validating safety systems, generating training data, and testing edge cases. This multi-functionality is achieved by creating a universal scenario generation system that can produce various traffic conditions from a single integrated platform

Inventive Principle:
Principle #6Universality (Multi-functionality)

Data Source

PatentEP3690754B1Method and device for creating traffic scenario with domain adaptation on virtual driving environment for testing, validating, and training autonomous vehicle
Publication Date: 2026.04.22 STRADVISION
  • EP3690754B1 patent drawingFigure 1
  • EP3690754B1 patent drawingFigure 2
  • EP3690754B1 patent drawingFigure 3

AI summary

A method for creating a traffic scenario in a virtual driving environment is provided. The method includes steps of: a traffic scenario-generating device, (a) on condition that driving data have been acquired which are created using previous traffic data corresponding to discrete traffic data extracted by a vision-based ADAS from a past driving video and detailed traffic data corresponding to sequential traffic data from sensors of data-collecting vehicles in a real driving environment, inputting the driving data into a scene analyzer to extract driving environment information and into a vehicle information extractor to extract vehicle status information on an ego vehicle, and generating sequential traffic logs according to a driving sequence; and (b) inputting the sequential traffic logs into a scenario augmentation network to augment the sequential traffic logs using critical events, and generate the traffic scenario, verifying the traffic scenario, and mapping the traffic scenario onto a traffic simulator.