Natural-Language Vehicle Testing for Flexible Sensor Data Capture
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing automated vehicle testing systems rely on strict, rule-based trigger conditions that can lead to missing necessary data collection and result in false positives/negatives, limiting the flexibility and accuracy of data collection.
Innovation Solution
Implementing a system that converts human-readable testing requirements and incident scenario descriptions into a coded file (RaC) using natural language, allowing for less strict data collection conditions and reducing false positives/negatives by using an ML model to evaluate sensor data against these requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If rule-based trigger conditions with explicit, deterministic, and clear definitions are used for sensor data collection, then data collection precision is improved, but data collection completeness deteriorates due to missing necessary data and false positives/negatives
Solution Approach 1:
The patent introduces an intermediary component (testing requirement management system) that translates natural language testing requirements into executable test cases. This intermediary layer bridges the gap between human-understandable requirements and machine-executable rules, allowing the system to capture nuanced testing scenarios that strict rule-based systems would miss, thereby improving data collection completeness while maintaining precision through structured translation processes
Solution Approach 2:
The system changes the parameter of requirement expression from rigid, structured formats to flexible natural language. By allowing testing requirements to be defined in natural language and then translating them into test cases, the system adapts to accommodate varied and complex testing scenarios without being constrained by predetermined rule structures, thus improving both completeness and precision
2Measurement precision
If strict conditions are imposed for sensor data collection, then false positives/negatives are reduced, but testing flexibility deteriorates
Solution Approach 1:
The patent implements a dynamic testing requirement management system where test cases are generated and updated based on natural language requirements that can be easily modified. The system dynamically adapts testing conditions by translating updated natural language requirements into new test cases, allowing testing flexibility to improve while maintaining accuracy through the structured translation process that preserves essential testing criteria
3Ease of operation
If natural language testing requirements are converted into coded files, then ease of operation is improved, but system complexity increases due to the conversion process
Solution Approach 1:
The system implements self-service automation where the testing requirement management system automatically translates natural language requirements into coded test cases without requiring manual intervention. This automated conversion process handles the complexity internally while presenting a simple natural language interface to users, thereby improving ease of operation while managing system complexity through automation rather than manual processes
Data Source
AI summary
Provided are a method, system, and device for automated vehicle testing. The method may include, generating a ticket, wherein the ticket comprises at least one natural language testing requirement and at least one natural language incident scenario description; determining whether collected sensor data from a vehicle matches the at least one natural language incident scenario description in the ticket; based on determining that the collected sensor data from the vehicle matches the at least one natural language incident scenario description: generating a requirements as code (RaC) file based on the ticket and collected sensor data from a vehicle; and evaluating a ML model based on the RaC file to determine whether the ML model achieves the testing requirement, wherein the ML model is used to implement a vehicle application.