Sensor Data Attribution for HD Map Compensation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for 3D road geometry modeling and sensor data analysis are resource-intensive and costly, making them impractical for high-definition map creation and updating, especially with the vast volume of crowd-sourced sensor data from vehicles, which requires efficient compensation mechanisms for data usage.
Innovation Solution
A system that aggregates sensor data from multiple sources, collapses vehicle paths onto road topology by travel direction, assigns observation counts, and projects these counts along predetermined distance units to form observation intervals, determining the value of sensor data based on interval lengths and counts, allowing for fair compensation without explicit traceability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If traditional methods for 3D road geometry modeling and sensor data analysis are used, then high-definition map creation and updating can be achieved, but the process becomes resource-intensive and costly
Solution Approach 1:
The patent segments the road network into discrete road segments and divides sensor data processing by individual data sources. Each road segment is processed independently with its own observation intervals, allowing parallel processing and reducing overall system complexity while maintaining high-definition map precision
Solution Approach 2:
The patent creates simplified representations of vehicle paths by collapsing them onto road topology to form observation intervals. These compressed data structures serve as copies that retain essential information for map creation while dramatically reducing the computational resources needed compared to processing raw sensor data
2Measurement precision
If manual or semi-automated analysis of large amounts of sensor data is performed, then data quality can be maintained, but the process is time-consuming and not practical for large-scale applications
Solution Approach 1:
The system automatically attributes sensor data to road segments and calculates observation intervals without human intervention. The automated process assigns observation counts to data sources and computes data values based on interval lengths, maintaining accuracy while enabling large-scale processing of crowd-sourced sensor data
Solution Approach 2:
The patent transforms raw sensor data into structured observation intervals with specific parameters (observation counts, interval lengths, data source identifiers). This parameterization enables efficient automated processing while preserving the essential information needed for accurate data analysis and valuation
3Adaptability or versatility
If compensation is provided to data sources for sensor data sharing, then data source participation is encouraged, but determining fair compensation becomes complex due to varying data volumes and qualities
Solution Approach 1:
The patent replaces complex manual negotiation and assessment mechanisms with an automated computational system. The system calculates compensation values by processing observation intervals and applying valuation algorithms, objectively determining fair compensation based on actual data contribution without human intervention
Solution Approach 2:
The patent introduces observation intervals as an intermediary structure between raw sensor data and compensation calculations. These intervals serve as a neutral representation that captures data contribution metrics, enabling transparent and fair compensation determination while simplifying the interaction between data sources and the map creation system
Data Source
AI summary
A method, apparatus and computer program product are provided for analyzing sensor data from roads of the road network to monitor sensor data usage, which can in turn be used for value estimation and compensation for data usage. Methods may include: aggregating observations and vehicle paths from sensor data from data sources; determining observation counts for each of the data sources; collapsing aggregated vehicle paths to road topology by travel direction; assigning the observation counts to a nearest road topology; projecting an observation count for each of the data sources along a predetermined distance unit of the road topology to form observation intervals, each observation interval including a respective observation count; iterating over the road topology using the observation intervals to establish observation intervals and their respective observation counts; and determining a value of the sensor data based on lengths of the observation intervals and their respective observation counts.