Road Defect Detection Fusion for Water-Filled Pothole Avoidance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing vehicle imaging systems struggle to accurately detect and classify potholes, especially when filled with water, due to limitations in visual contrast, radar scattering, and operational range, posing safety risks from unexpected impacts and loss of vehicle control.
Innovation Solution
A vehicular driving assist system that combines data from multiple sensors (cameras, radar, lidar, and HD maps) to detect and classify potholes based on size, shape, and depth, using a fusion algorithm to enhance detection accuracy and reliability, and trigger appropriate alerts or vehicle responses.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If visual sensors (cameras) are used to detect potholes, then the system can identify road defects, but detection accuracy deteriorates when potholes are filled with water due to loss of visual contrast
Solution Approach 1:
The patent combines multiple sensor types (cameras, radar, lidar, ultrasonic sensors) into a unified detection system. Each sensor type compensates for the weaknesses of others - for example, radar and lidar can detect water-filled potholes through the water surface, while ultrasonic sensors provide depth information that cameras cannot obtain visually.
Solution Approach 2:
The system employs multi-functional sensors that can operate under various road conditions. The radar system, for instance, can detect both dry and water-filled potholes, while the lidar provides both surface mapping and depth measurement capabilities, making the overall system universally effective across different adverse conditions.
2Reliability
If radar sensors are used to detect potholes, then the system can operate in adverse weather conditions, but detection precision deteriorates due to radar scattering from water surfaces
Solution Approach 1:
The system uses lidar as an intermediary sensor to provide initial pothole location and surface topology information. This lidar data serves as a reference that helps the radar system distinguish between actual potholes and false targets caused by water surface scattering, thereby improving radar detection precision in adverse weather.
Solution Approach 2:
The system implements feedback mechanisms where detection results from multiple sensors are continuously cross-validated. When radar detects a potential pothole, the system uses feedback from lidar and ultrasonic sensors to verify the detection and eliminate false positives caused by water surface scattering.
3Measurement precision
If ultrasonic sensors are used to detect potholes, then the system can measure pothole depth, but the operational range is limited and cannot detect distant potholes
Solution Approach 1:
The system transitions from one-dimensional depth measurement (ultrasonic) to three-dimensional spatial mapping by integrating lidar data. The lidar provides comprehensive 3D surface topology information that covers long distances, while ultrasonic sensors provide detailed depth verification for nearby detected potholes, creating a multi-layered detection approach.
Solution Approach 2:
The system uses lidar and radar for preliminary long-range pothole detection and location identification. Once potential potholes are identified at a distance, ultrasonic sensors are deployed for preliminary depth measurement as the vehicle approaches, ensuring continuous monitoring across all ranges.
4Reliability
If multiple sensors are combined to improve detection accuracy, then detection reliability in adverse conditions improves, but system complexity increases
Solution Approach 1:
The detection system is segmented into specialized subsystems: a long-range detection subsystem (lidar and radar), a medium-range verification subsystem (cameras and radar), and a short-range measurement subsystem (ultrasonic sensors). Each subsystem is optimized for specific detection tasks and operational ranges, reducing overall system complexity through functional segmentation.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Improves pothole detection and classification accuracy, especially in adverse conditions, enabling proactive alerts and vehicle adjustments to mitigate potential hazards, enhancing safety and comfort.
Implementation Method 1
A vehicular driving assist system includes a sensor disposed at a vehicle equipped with the vehicular driving assist system. The sensor senses exterior of the vehicle and is operable to capture sensor data.
Data Source
AI summary
A vehicular driving assist system includes a sensor disposed at a vehicle and operable to capture sensor data. The vehicular driving assist system, responsive to processing at an ECU of sensor data captured by the sensor, detects a pothole in ahead of the vehicle. The vehicular driving assist system, responsive to detecting the pothole, determines severity of the detected pothole based at least in part on geographic-linked data that includes at least one of size, depth or shape of the detected pothole. At least partially based on the severity of the pothole, at least one of (i) a driver of the vehicle is alerted, (ii) steering of the vehicle is controlled to mitigate a wheel of the vehicle traveling over the detected pothole and (iii) braking of the vehicle is controlled to mitigate the wheel of the vehicle traveling over with the detected pothole.