Siamese Network for At-Risk Road Infrastructure Identification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current technologies face challenges in predicting and identifying at-risk road infrastructure changes within a road network, relying on collected data from specific locations without a predictive framework to focus map update efforts effectively.
Innovation Solution
A method utilizing a Siamese network to analyze sensor data against historical map data, differentiating visual similarity from map update similarity, and calculating a change likelihood score to identify at-risk infrastructure by comparing sensor data to both updated and non-updated map data, thereby prioritizing map update requests.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If map data is updated using crowd sourced probe data from vehicles, then map coverage can be expanded, but the accuracy and reliability of map updates deteriorates due to data variability
Solution Approach 1:
The patent introduces an intermediary verification system that compares crowd sourced probe data against historical map data and sensor data from multiple sources. This intermediary layer filters and validates the variable probe data before incorporating it into map updates, thereby maintaining accuracy while expanding coverage.
Solution Approach 2:
The system implements feedback mechanisms where map update decisions are continuously refined based on comparison results between probe data, historical map data, and sensor data. The feedback loop allows the system to learn from data variability and improve validation criteria over time, maintaining accuracy as coverage expands.
2Measurement precision
If map updates are performed at all locations, then map data accuracy is maintained, but the time and resources required for updates increase significantly
Solution Approach 1:
The patent applies local quality by focusing map update efforts only on specific locations where changes are detected through the comparison system. Rather than uniformly updating all map data, the system identifies and prioritizes locations with detected infrastructure changes, thereby maintaining accuracy while reducing overall update time and resources.
Solution Approach 2:
The patent segments the map update process into two distinct phases: a screening phase that quickly identifies potential changes using the comparison system, and a detailed update phase that focuses resources only on identified change locations. This segmentation eliminates unnecessary updates to unchanged areas, significantly reducing time loss.
3Reliability
If historical map data with update histories is used for comparison, then the ability to identify at-risk infrastructure improves, but the complexity of data processing increases
Solution Approach 1:
The patent extracts and utilizes specifically the update history component from historical map data, rather than processing all historical data. By focusing only on the relevant update history information and comparing it with current sensor data, the system achieves reliable at-risk infrastructure identification while minimizing processing complexity.
Data Source
AI summary
A method, apparatus and computer program product are provided for predicting the likelihood that road infrastructure has changed. Methods may include: receiving sensor data from a sensor network of a first environment; identifying first map data of an environment, where the first map data is identified based on visual similarity to the first environment, where the first map data includes a history of map updates; identifying second map data of an environment, where the second map data is identified based on visual similarity to the first environment; analyzing the sensor data against the first map data and the second map data to establish correspondence between the sensor data and the first map data and between the sensor data and the second map data; and identifying the first environment as a location of at-risk infrastructure based on the sensor data corresponding to the first map data.