Autonomous Vehicle Software Verification via Simulation Precertification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The widespread adoption of autonomous vehicles on public roads is hindered by the need for extensive real-world testing and a convincing safety record, as well as limitations in monetization strategies for autonomous technology, which rely on gradual progression of autonomy features in vehicles.
Innovation Solution
An on-demand transportation management system that utilizes risk regression and trip classification techniques to dynamically assess and manage autonomous vehicle operations, integrating fully autonomous vehicles with human-driven vehicles, and employs simulation-based precertification of software releases to ensure safety and efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If extensive real-world testing is conducted to establish safety record, then reliability of autonomous vehicles improves, but loss of time and productivity increases
Solution Approach 1:
The patent applies simulation-based precertification of software releases before real-world deployment. Multiple software versions are pre-tested in virtual environments using simulation engines that model various driving scenarios, weather conditions, and road types. This preliminary action in silico allows the system to verify software safety and performance before committing to extensive real-world testing, thereby reducing the time and productivity loss associated with sequential testing while maintaining reliability standards.
2Measurement precision
If software versions are extensively tested in real-world conditions, then measurement precision of safety metrics improves, but loss of time increases
Solution Approach 1:
The patent creates virtual copies of real-world driving environments through simulation engines. These digital twins replicate road geometries, weather conditions, traffic patterns, and sensor behaviors. By testing software versions in these copied environments, the system achieves high measurement precision for safety metrics without the time costs of physical testing. The simulation-based approach allows parallel testing of multiple software versions across diverse scenarios simultaneously, accelerating verification while maintaining metric accuracy.
Solution Approach 2:
The patent transitions testing from the physical dimension to the virtual/digital dimension. Instead of testing software exclusively in physical vehicles on real roads, the system employs simulation environments that exist in a different dimensional space. This dimensional shift enables faster iteration and testing while preserving the essential characteristics needed for safety metric validation, thereby reducing time loss without sacrificing measurement precision.
3Productivity
If autonomous vehicle operations are expanded rapidly, then productivity increases, but reliability may deteriorate due to insufficient testing
Solution Approach 1:
The patent implements simulation-based precertification as a preliminary gate before real-world deployment. Software versions undergo comprehensive virtual testing across multiple simulated scenarios including edge cases, adverse weather, and complex traffic situations. Only software that passes these preliminary simulation tests is deployed to autonomous vehicles. This preliminary action enables rapid operational expansion while maintaining reliability standards, as the simulation filter ensures only adequately tested software reaches the fleet.
Solution Approach 2:
The patent introduces simulation engines as an intermediary between software development and real-world deployment. Rather than directly deploying untested software to vehicles, the simulation environment acts as a mediator that validates software performance first. This intermediary layer enables rapid scaling of autonomous operations by providing a safe, accelerated testing pathway that maintains reliability without requiring proportional increases in real-world testing time.
Data Source
AI summary
An autonomous vehicle software management system can distribute AV software versions to safety-driven autonomous vehicles (SDAVs) operating within a given region. The system can receive log data from the SDAVs indicating any trip anomalies of the SDAVs while executing the AV software version. When a predetermined safety standard has been met based on the log data, the system can verify the AV software version for execution on fully autonomous vehicles (FAVs) operating within the given region.