Vehicle Localization Validation Using Independent Point Cloud Metrics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing autonomous vehicle localization systems face challenges in ensuring functional safety, particularly in validating the accuracy of localization data across varying environmental conditions and hardware limitations.
Innovation Solution
The implementation of a localization functional safety system that iteratively processes point cloud data to calculate a transformation between the point cloud and a map, using a rotation matrix and transformation vector to determine the vehicle's pose. This system applies a transform to the source point cloud and calculates metrics to validate the localization within predetermined thresholds, utilizing hardware and software independent from the primary localization computation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a localization system uses iterative point cloud processing to calculate transformation and validate pose accuracy, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The localization system is divided into separate functional modules: a first processor that calculates the primary transformation and pose, and a second processor that performs validation by calculating alternative metrics. This segmentation allows each module to focus on specific computational tasks, improving overall measurement precision while managing system complexity through modular design.
Solution Approach 2:
A validation processor acts as an intermediary between the primary localization computation and the final pose output. This intermediary independently verifies the transformation calculations by computing alternative metrics and comparing results, thereby enhancing measurement precision without requiring the entire system to operate at maximum complexity.
2Reliability
If dual independent processors are used to validate localization metrics, then reliability is improved, but use of energy increases
Solution Approach 1:
The validation processor performs partial verification by calculating only the necessary alternative metrics required to validate the primary localization results, rather than performing complete redundant localization computations. This partial action approach maintains functional safety and reliability while significantly reducing the energy consumption compared to full duplicate processing.
Solution Approach 2:
The system changes computational parameters by having the validation processor use different calculation methods or simplified algorithms compared to the primary processor. This allows the validation function to achieve reliable verification with reduced computational intensity, thereby lowering energy consumption while maintaining safety integrity.
3Measurement precision
If comprehensive localization validation is performed, then measurement precision is improved, but loss of time occurs
Solution Approach 1:
The system performs preliminary validation checks during the primary localization computation process, rather than conducting complete separate validation after pose determination. By integrating validation steps into the main computational flow, the system achieves comprehensive measurement precision without the time loss of sequential processing.
Solution Approach 2:
The validation processor operates continuously alongside the primary localization system, performing verification computations in parallel rather than sequentially. This continuous action ensures that localization precision is maintained through ongoing validation without interrupting or delaying the primary navigation functions.
Data Source
AI summary
Provided are methods for localization functional safety, which can include systems, methods, and computer program products are also provided. In examples, a method includes applying a transform to a source point cloud and calculating a second metric based on the application of the transform to the source point cloud and a map at a higher ASIL level. A first metric is determined based on a localization function that executes at a lower ASIL level. A deviation between a first metric and the second metric, is determined wherein the vehicle localization is validated when the deviation is less than a predetermined threshold.