Autonomous Vehicle Scenario Matching for Adaptive Maneuver Constraints
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face challenges in navigating scenarios without known matches, leading to potential infractions due to the lack of effective manoeuvring constraints.
Innovation Solution
A computer-implemented method involving scenario encoding and comparison using variational autoencoders, dynamic parameter simulations, and weighted matchers to determine and control manoeuvring constraints, including minimal risk manoeuvres when no exact match is found.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the AV uses a database of known scenarios with predefined travel envelopes, then navigation safety is improved for known scenarios, but the system cannot handle novel scenarios without known matches
Solution Approach 1:
The patent introduces an intermediary system that computes manoeuvring constraints by comparing current sensor data with a database of known scenarios using distance metrics. When no exact match is found, the system interpolates constraints from similar scenarios, acting as a mediator between known scenario data and novel situation handling.
Solution Approach 2:
The system pre-computes and stores manoeuvring constraints for known scenarios in a database before deployment. During operation, these pre-computed constraints are quickly retrieved and adapted through interpolation when needed, avoiding real-time complex calculations and enabling rapid response to novel scenarios.
2Reliability
If the AV performs minimal risk manoeuvres for unknown scenarios, then safety is improved, but productivity and efficiency deteriorate due to conservative behavior
Solution Approach 1:
The system dynamically adjusts the degree of conservatism in manoeuvring based on the computed distance to known scenarios. When distance is small (similar scenarios), the system allows more aggressive manoeuvres. When distance is large (novel scenarios), it defaults to minimal risk manoeuvres. This continuous parameter adjustment optimizes both safety and efficiency.
Solution Approach 2:
The manoeuvring constraint system is dynamic rather than static. It continuously adapts constraints based on real-time comparison with known scenarios, allowing the AV to transition between conservative and aggressive behaviours depending on environmental familiarity, thereby balancing safety and productivity.
3Adaptability or versatility
If the AV interpolates manoeuvring constraints for scenarios between known matches, then adaptability is improved, but the complexity of the control system increases
Solution Approach 1:
The system creates simplified copies of known scenario constraints and interpolates between them for novel scenarios. Rather than developing entirely new constraints from scratch, it replicates and combines existing constraint patterns, reducing the complexity of the control system while maintaining adaptability.
4Measurement precision
If the AV uses multiple weighted matchers for scenario comparison, then measurement precision is improved, but computational complexity and time increase
Solution Approach 1:
The system uses multiple weighted matchers but applies them selectively based on the situation. Not all matchers are always activated or weighted equally - the system adjusts the level of matching scrutiny based on scenario characteristics, achieving sufficient precision without always incurring maximum computational cost.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The present invention relates to a computer-implemented method of controlling an autonomous vehicle. The method comprises: receiving current sensor data defining a current scenario; determining a distance between the current scenario and a closest scenario from a database of known scenarios; and controlling the autonomous vehicle by: manoeuvring the autonomous vehicle within a manoeuvring constraint associated with the closest scenario when the distance is below a first threshold; performing a minimal risk manoeuvre when the distance is above a second threshold, wherein the second threshold is greater than the first threshold; and interpolating the manoeuvring constraint associated with the closest scenario when the distance is between the first threshold and the second threshold, and manoeuvring the autonomous vehicle within the interpolated manoeuvring constraint.