Autonomous Vehicle Hazard Detection via Sensor Fusion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing navigation systems fail to accurately update maps in real-time to account for weather-based hazards such as rain, fog, and slippery roads, leading to increased travel times and potential accidents for autonomous vehicles.
Innovation Solution
A system and method that combines vehicle sensor data and image data using a Convolutional Neural Network model to generate hazard polygons, which are then used to update map data in real-time, providing better route guidance and warnings for autonomous vehicles.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If real-time map updating using vehicle sensor data and image data is implemented, then detection accuracy of weather-based hazards is improved, but system complexity increases
Solution Approach 1:
The patent combines vehicle sensor data (LiDAR, cameras, radar) with map data and image data from multiple sources into a unified processing system. This merging of diverse data streams enables comprehensive hazard detection while sharing processing infrastructure to manage system complexity.
Solution Approach 2:
The system introduces an intermediary processing layer that receives raw sensor data, processes it through neural networks, and generates standardized hazard polygons. This intermediary layer abstracts the complexity of multi-source data integration while maintaining high detection accuracy through specialized processing pipelines.
2Reliability
If multiple data sources are integrated for hazard detection, then reliability of hazard warning is improved, but processing time increases
Solution Approach 1:
The system pre-processes sensor data and maintains ready-to-use hazard polygons for common weather conditions. When hazards are detected, pre-configured processing templates are applied immediately, reducing real-time processing time while maintaining reliable multi-source validation.
Solution Approach 2:
The processing system is divided into independent modules that handle different data sources (sensor data processing, image data processing, map data integration) separately. These segmented modules can process data in parallel and combine results, reducing overall processing time while maintaining comprehensive reliability through multiple validation points.
3Reliability
If hazard polygons are generated and updated in real-time, then navigation safety is improved, but computational resources required increase
Solution Approach 1:
The system updates hazard polygons at periodic intervals based on vehicle position changes and environmental conditions rather than continuously. This periodic updating reduces computational resource consumption while maintaining navigation safety by ensuring hazard information is current at regular checkpoints during vehicle operation.
Data Source
AI summary
The disclosure provides a method, a system, and a computer program product for updating map data. The method comprises obtaining vehicle sensor data associated with first spatial data and first temporal data related to one or more hazard-based event. The method may further include obtaining image data associated with second spatial data and second temporal data related to one or more hazard-based event. The method may further include combining the first spatial data with the second spatial data based on a match between the first spatial data and the second spatial data and between the first temporal data and the second temporal data respectively. The method may further include updating the map data based on the combining.