Driver Assistance Validation via Critical Range Testing

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current methods for validating driver assistance systems in vehicles are limited, as they primarily rely on stochastic simulations and real accident data, neglecting critical scenarios where no accidents occurred or were prevented, thus missing potential for system development.

Innovation Solution

The method involves defining and altering test parameters to focus on critical ranges where driver assistance systems intervene, using both real and virtual testing environments, and sensor data to generate new test scenarios, allowing for the identification of hidden defects and interactions between systems.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If stochastic simulation methods are used to generate test scenarios, then a large number of diverse scenarios can be generated, but the scenarios are generated solely by random chance and do not focus on critical ranges where driver assistance systems intervene

Engineering Contradiction:

Improvediversity of test scenariosVSAvoidfocus on critical ranges

Core Design Contradiction:

Adaptability or versatilityVSMeasurement precision

Solution Approach 1:

The method performs preliminary analysis of measurement data to identify critical ranges before generating test scenarios. By pre-determining which parameter ranges are critical for driver assistance system intervention, the system can then focus scenario generation on these identified ranges rather than relying solely on random stochastic generation.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The method uses feedback from measurement data and previous test results to iteratively refine and adjust test scenario generation. Measurement data from real driving situations is analyzed to identify patterns and critical ranges, which then feed back into the scenario generation process to create more targeted and relevant test cases.

Inventive Principle:

Principle #23Feedback

2Reliability

If only real accident data is used for simulation, then scenarios based on actual accidents can be evaluated, but scenarios for successfully prevented accidents or near misses are discarded despite their value for testing critical intervention ranges

Engineering Contradiction:

Improveaccuracy of accident evaluationVSAvoiddiscarded measurement data

Core Design Contradiction:

ReliabilityVSLoss of information

Solution Approach 1:

The method converts previously discarded measurement data from non-accident situations (successfully prevented accidents, near misses) into valuable test resources. By recognizing that these situations represent critical ranges where driver assistance systems successfully intervened or could have intervened, the system transforms what was considered useless data into beneficial test scenarios for validating system performance in critical situations.

Inventive Principle:

Principle #22Blessing in disguise (Convert harm into benefit)

Solution Approach 2:

The method creates a universal approach that handles multiple types of driving situations uniformly - both actual accidents and successfully prevented accidents/near misses are treated as valuable sources for test scenario generation. This multi-functional use of measurement data allows the system to extract test scenarios from any driving situation, maximizing the utility of available data.

Inventive Principle:

Principle #6Universality (Multi-functionality)

3Ease of operation

If test parameters are not systematically altered to displace them within critical ranges, then tests can be performed with simple parameters, but hidden defects in the driver assistance system may remain undetected

Engineering Contradiction:

Improvesimplicity of test executionVSAvoiddetection of hidden defects

Core Design Contradiction:

Ease of operationVSReliability

Solution Approach 1:

The method dynamically adjusts test parameters by systematically altering them to displace into critical ranges identified from measurement data. Rather than using static, simple test parameters, the system dynamically modifies parameters such as distance to leading vehicle, relative velocity, and acceleration to create test scenarios that actively probe the critical intervention ranges of the driver assistance system, thereby revealing hidden defects.

Inventive Principle:

Principle #15Dynamics

Data Source

PatentUS9937930B2Virtual test optimization for driver assistance systems

Publication Date: 2018.04.10 AVL LIST GMBH

AI summary

A method for validating a driver assistance system (3) of a vehicle, wherein tests (T) defined by test parameters (P) are carried out for a predetermined test scenario (4), during a first test (T(n)) at least on test parameter (P) is determined, and to generate a second test (T(n+1)) the first test (T(n)) is altered in order to displace the first test parameter (P) within a critical range (7) assigned to it.