Vehicle Safety Function Validation Using Activation Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for validating vehicle safety systems require extensive driving and data collection, necessitating a large number of vehicles and long development times, which is inefficient and costly, especially for features like forward-looking collision mitigation with brake-intervention, due to the need for representative data across various driving scenarios and high computing power for simulations.
Innovation Solution
The approach involves collecting activation data from software programs running on vehicle controllers without actual actuation, using a classifier to differentiate between beneficial and false activations, and selecting the optimal software based on benefit-to-false-activation ratio, allowing for reduced data collection and faster development by leveraging existing vehicles and simulations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If extensive driving and data collection is performed to validate safety systems, then validation reliability is improved, but development time and cost increase significantly
Solution Approach 1:
The patent applies preliminary action by collecting and storing validation data during normal vehicle operation before formal validation is needed. Data is continuously recorded from sensors and vehicle systems during regular driving, so when validation requirements arise, the data is already available for immediate analysis, eliminating the need for extended dedicated testing periods.
Solution Approach 2:
The patent uses copying by creating virtual representations of validation scenarios through simulation. Instead of physically driving vehicles through every possible test scenario, the system uses recorded real-world data to create virtual test environments that can be repeatedly analyzed and validated without additional physical testing, significantly reducing time and resource requirements.
2Reliability
If a large number of vehicles are deployed for data collection, then data coverage and representativeness are improved, but resource requirements and costs increase
Solution Approach 1:
The patent applies universality by designing a data collection system that serves multiple functions simultaneously. The same vehicle sensors and data recording infrastructure used for normal vehicle operation are also used for validation data collection, eliminating the need for separate dedicated test vehicles. This multi-functional approach allows data to be collected from regular production vehicles without requiring additional test fleets.
Solution Approach 2:
The patent merges the validation data collection function with normal vehicle operation. Instead of separating test vehicles from production vehicles, the system combines both functions into the same vehicle fleet, using the same hardware and software infrastructure for both regular operation and validation data gathering, thereby reducing the total number of vehicles needed.
3Reliability
If all recorded data is re-simulated to guarantee software fulfillment, then false activation rate is reduced, but computing power requirements and processing time increase
Solution Approach 1:
The patent extracts only the critical and representative portions of recorded data for simulation analysis rather than processing all data. The system identifies key validation scenarios and extracts relevant data subsets that are sufficient to validate safety functions, eliminating the need to simulate entire datasets and significantly reducing computing power requirements while maintaining validation effectiveness.
Solution Approach 2:
The patent applies partial action by performing simulation on a selected subset of recorded data that is sufficient for validation purposes. Instead of exhaustively simulating all recorded miles and scenarios, the system identifies representative samples that provide adequate statistical confidence for validation, reducing computational burden while maintaining reliability standards.
Data Source
AI summary
Methods and systems of validating a vehicle safety function. In one implementation, a plurality of vehicle safety function software programs are developed. A first one of the software programs is installed in a memory of a vehicle controller. The program is run on the controller without delivering at least one vehicle actuator output of the controller generated as a result of running the program. Activation data for the safety function is recorded and the steps of installing, running, and recording with a second one of the plurality of vehicle safety function software programs are repeated.