Wheel Sensor Road Unevenness Detection With Edge Characterization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for detecting and characterizing road unevenness, such as potholes, using sensors like lidar, radar, or camera systems are unreliable due to susceptibility to false positives and negatives, high computing resource consumption, and failure to meet safety standards like ASIL-D, and lack region-specific data for creating hazard maps.
Innovation Solution
A method and device utilizing wheel speed sensors and wheel-specific acceleration sensors to determine and characterize road unevenness by analyzing edge shapes, with an arithmetic unit processing sensor data to detect and classify road conditions, including edge steepness, type, and properties like depth and length, using threshold values and machine learning models for precise detection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Difficulty of detecting and measuring
If lidar, radar, or camera sensors are used to detect road unevenness, then detection capability is improved, but reliability and safety standard compliance deteriorate
Solution Approach 1:
The patent replaces optical sensors (lidar, radar, camera) with mechanical sensors (wheel speed sensors and acceleration sensors) that are already integrated into the vehicle. These mechanical sensors directly measure wheel behavior when encountering road unevenness, providing ASIL-D compliant reliable data without the false positive/negative issues of optical systems.
Solution Approach 2:
The invention utilizes sensors that are already present in the vehicle (wheel speed sensors and acceleration sensors) for their primary function, and repurposes them to also detect road unevenness. This self-service approach eliminates the need for additional specialized sensors while maintaining reliability standards.
2Measurement precision
If machine learning algorithms are used to detect road unevenness, then detection accuracy is improved, but computing time consumption increases
Solution Approach 1:
The patent extracts only the essential features needed for road unevenness detection from the sensor data (wheel speed changes and acceleration values), avoiding the need for complex full-image or full-point-cloud processing. This extraction approach maintains detection accuracy while dramatically reducing computing time requirements.
Solution Approach 2:
The invention segments the detection task into specific measurable parameters (wheel speed differential, acceleration magnitude, duration of event) rather than attempting to analyze complete sensor datasets. This segmentation enables faster processing while maintaining precision for the specific task of detecting potholes and road unevenness.
3Adaptability or versatility
If general sensors are used for road detection, then sensor availability is improved, but measurement precision and reliability deteriorate
Solution Approach 1:
The patent makes existing wheel sensors serve multiple functions: their primary function for vehicle speed monitoring and their secondary function for road unevenness detection. This multi-functionality ensures high sensor availability across all vehicles while achieving precise road characterization through clever signal analysis of the existing sensor data.
Data Source
AI summary
A method for determining and characterizing road unevenness of a roadway. Sensor data are generated by at least one wheel speed sensor and/or at least one wheel-individual acceleration sensor of a motor vehicle travelling the roadway. The road unevenness is determined and characterized by an arithmetic unit using the sensor data generated, thereby determining an edge type of the road unevenness.