Autonomous Vehicle Simulation Using Log-Based Object Trajectories
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Creating realistic driving simulations that accurately reflect real-world scenarios is challenging due to the difficulty in replicating the behavior of simulated agents, which often deviate from their real-world counterparts.
Innovation Solution
The use of hybrid log-based simulations, where log data from real-world vehicle operations is used to instantiate simulated vehicles and environments, allowing for the recreation of real-world scenarios and behaviors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional driving simulations are used, then testing can be performed in a controlled environment, but the simulated agents exhibit behavior that deviates from real-world counterparts
Solution Approach 1:
The patent uses recorded log data from real-world autonomous vehicle operations to create simulated scenarios. Instead of modeling complex real-world physics and behaviors, the system directly copies actual sensor data, object trajectories, and environmental conditions from logged drives to construct simulation scenarios, ensuring simulated agents behave exactly as they did in reality
Solution Approach 2:
The system performs preliminary data collection and processing by recording extensive log data from real vehicle operations before simulation is needed. This pre-captured data including sensor readings, object detections, and vehicle states is stored and can be quickly reused to generate multiple simulation scenarios without requiring complex real-time modeling
2Reliability
If real-world testing is performed, then accurate vehicle behavior can be observed, but safety concerns and practical limitations arise
Solution Approach 1:
The patent introduces a simulated environment as an intermediary between real-world data and vehicle testing. Real log data is used to construct virtual scenarios where simulated vehicles and objects interact according to recorded real-world behaviors, allowing safe testing of edge cases and dangerous situations that would be impractical to recreate in the physical world
Solution Approach 2:
The system copies actual dangerous or rare real-world scenarios from log data into the simulation environment, preserving the authenticity of hazardous situations without exposing physical vehicles or personnel to actual risk. This includes replicating pedestrian crossings, vehicle interactions, and environmental conditions from recorded incidents
Data Source
AI summary
Techniques for determining a response of a simulated vehicle to a simulated object in a simulation are discussed herein. Log data captured by a physical vehicle in an environment can be received. Object data representing an object in the log data can be used to instantiate a simulated object in a simulation to determine a response of a simulated vehicle to the simulated object. Additionally, one or more trajectory segments in a trajectory library representing the log data can be determined and instantiated as a trajectory of the simulated object in order to increase the accuracy and realism of the simulation.