Shadow Driving Risk Evaluation for Self-Driving Safety
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current self-driving safety evaluation methods rely on simulated systems, which fail to accurately reflect real-world scenarios, leading to significant deviations in safety assessments and high costs due to the need for extensive data collection, such as 5 million to 6 million kilometers of travel data.
Innovation Solution
A method that calculates risk values in both shadow driving and self-driving modes based on real-time and preset routes, using parameters like speed, time-to-collision, and distance to traffic signals, allowing for more accurate safety evaluations by re-planning self-driving routes based on manual driving data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If safety evaluation is performed using a simulated self-driving system, then the cost cycle for mass production is reduced, but the deviation from real self-driving safety increases
Solution Approach 1:
The patent performs preliminary safety evaluations using simulated self-driving systems before mass production, allowing manufacturers to identify and address safety issues early in the development phase. This preliminary action reduces the overall cost cycle by preventing costly recalls and modifications after deployment, while maintaining acceptable accuracy through iterative simulation refinement.
Solution Approach 2:
The patent introduces an intermediary layer of shadow driving data collection and multi-mode evaluation (combining simulated, shadow, and real driving data) that bridges the gap between simulation and real-world performance. This intermediary approach gradually improves measurement precision while managing costs through phased data collection and evaluation.
2Measurement precision
If extensive travelling data (5-6 million kilometers) is collected for safety evaluation, then the accuracy of safety assessment improves, but the cost for automobile manufacturers increases significantly
Solution Approach 1:
The patent segments the data collection and evaluation process into multiple phases and modes: simulated driving evaluation, shadow driving data collection, and real driving validation. Each segment contributes differently to safety assessment, allowing manufacturers to achieve comprehensive evaluation without requiring all vehicles to accumulate 5-6 million kilometers of real-world data, thereby reducing overall costs while maintaining accuracy.
Solution Approach 2:
The patent uses shadow driving mode as a copy of real driving conditions, where the self-driving system operates in parallel with manual driving to collect comparable data without the full cost and time requirements of actual mass deployment. This copying approach provides sufficient data for accurate safety assessment at reduced cost.
3Ease of operation
If the self-driving system follows preset routes, then the evaluation process is simplified, but the ability to reflect real-world driving scenarios decreases
Solution Approach 1:
The patent implements dynamic route adjustment where the self-driving system can switch between preset routes for standardized evaluation and real-time route planning for shadow driving based on actual road conditions and driver behavior. This dynamic approach maintains evaluation simplicity through preset routes while improving adaptability when shadow driving requires response to real-world scenarios.
Solution Approach 2:
The patent changes the route parameter from fixed preset routes to flexible real-time routes during shadow driving mode. By adjusting the route planning parameter based on driving mode (simulated vs. shadow vs. real), the system maintains simplicity where needed while adapting to real-world complexity when necessary for accurate safety evaluation.
Data Source
AI summary
A self-driving safety evaluation method, including determining RB based on a risk value of a vehicle in a shadow driving mode in a first measurement unit, where RB is a risk value of the vehicle in the shadow driving mode in a plurality of measurement units, and determining RC based on a risk value of the vehicle in a self-driving mode based on a preset route in the first measurement unit, where RC is a risk value of the vehicle in the self-driving mode based on the preset route in the measurement units, where RB and RC are used to determine whether safety of the vehicle in the self-driving mode meets a requirement.