Autonomous Vehicle Simulation with Modified Agents for Rare Road Users
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for testing autonomous vehicle software struggle to validate safety and effectiveness, especially in rare or outlier situations, such as interactions with less common road users like motorcyclists, due to limited real-world data examples, leading to insufficient testing of behaviors like changing lanes or merging near these users.
Innovation Solution
The method involves running simulations using log data from autonomous vehicle operations, where common road users are modified to simulate less common users, and interactive agents are introduced to respond to the vehicle's behavior, allowing for the analysis of interactions like collisions or near collisions, and flagging simulations for further review based on these interactions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If real-world data is used for testing autonomous vehicle software, then the testing reflects actual driving conditions, but the limited availability of rare situation data (e.g., interactions with motorcyclists) results in insufficient testing coverage
Solution Approach 1:
The patent creates synthetic copies of real-world driving scenarios by modifying log data. It generates modified versions of original driving situations where road users are transformed into different types (e.g., converting car interactions into motorcycle interactions), thereby creating test cases for rare situations without requiring actual real-world data of those specific scenarios.
Solution Approach 2:
The system changes parameters of existing log data to create new test scenarios. By modifying characteristics such as road user type, behavior patterns, and interaction outcomes in the log data, the system generates diverse test cases that cover rare situations while maintaining the structural integrity of original driving scenarios.
2Adaptability or versatility
If simulations are run with modified agents to test rare situations, then testing coverage for outlier cases is improved, but the complexity of managing modified and interactive agents increases
Solution Approach 1:
The patent segments the simulation system into distinct components: original log data processing, modified agent generation, and interactive agent implementation. This segmentation allows each component to be developed and managed independently, reducing overall system complexity while enabling comprehensive testing of rare situations through coordinated operation of these modular elements.
3Reliability
If interactive agents are introduced to replace modified agents during simulation, then realistic response to vehicle behaviors is achieved, but the computational resources and processing time required increase
Solution Approach 1:
The system performs preliminary actions by pre-processing log data and creating modified agents before running the actual simulation. This preparation work includes identifying scenarios that require interactive agents and pre-configuring them, so that during simulation execution, the replacement and interaction can occur more efficiently, reducing overall processing time.
Data Source
AI summary
The disclosure relates to testing software for operating an autonomous vehicle. In one instance, a simulation may be run using log data collected by a vehicle operating in an autonomous driving mode. The simulation may be run using the software to control a simulated vehicle and by modifying a characteristic of an agent identified in the log data. During the running of the simulation, that a first type of interaction between the first simulated vehicle and the modified agent will occur may be determined. In response to determining that the particular type of interaction will occur, the modified agent may be replaced by a interactive agent that simulates a road user corresponding to the modified agent that is capable of responding to actions performed by simulated vehicles. That the particular type of interaction between the simulated vehicle and the interactive agent has occurred in the simulation may be determined.