Map Data Collection Prediction for Road Sections
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing techniques for generating accurate road maps for automated vehicle-driving systems often fail to collect sufficient data in a predetermined period for certain road sections, leading to incomplete map updates.
Innovation Solution
An apparatus and method that predict the number of map-generating data pieces needed for under-collected road sections by analyzing traffic volume history and environmental conditions, instructing data collection devices to gather additional data when necessary to reach target numbers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If data collection relies on natural traffic flow without intervention, then device complexity and operational simplicity are maintained, but the quantity of map-generating data collected for certain road sections fails to reach target numbers within predetermined periods
Solution Approach 1:
The system performs preliminary analysis of traffic volume history and environmental conditions to predict future data collection outcomes before actually collecting data. This allows the management device to proactively identify road sections that will fail to meet target numbers and intervene with collection instructions in advance, ensuring target numbers are achieved within predetermined periods rather than discovering deficiencies after the fact.
Solution Approach 2:
The system establishes a feedback loop where collection results (actual numbers of map-generating data pieces) are compared against target numbers, and the difference feeds back into the prediction model. The management device uses this feedback to continuously refine traffic volume predictions and adjust collection instructions, ensuring that target numbers are consistently met across different road sections and time periods.
2Quantity of substance
If the system collects data from all road sections uniformly, then data collection simplicity is maintained, but resources are wasted on road sections that already exceed target numbers while insufficient data is gathered from under-collected sections
Solution Approach 1:
The system applies different data collection strategies to different road sections based on their specific characteristics. Road sections are categorized as under-collected, adequately collected, or over-collected based on comparisons between actual and target numbers. Collection instructions are selectively applied only to under-collected sections, with parameters tailored to each section's predicted traffic volume and data needs, optimizing resource allocation across the entire road network.
Solution Approach 2:
The system dynamically changes collection parameters (such as collection probability, interval, or duration) based on predicted traffic volume and the gap between actual and target numbers. For road sections with high predicted traffic volume, the system may use lower collection probabilities to avoid redundancy, while for sections with low predicted volume, it may increase collection intensity. This parameter adaptation ensures efficient resource use while meeting target numbers.
Data Source
AI summary
An apparatus for collecting map-generating data includes a processor configured to count, for each of road sections, the number of pieces of map-generating data received in a first period from one of at least one vehicle, identify one of the road sections for which the number of pieces of map-generating data received in the first period does not reach a target number for the one of the road sections, and predict, for the identified road section, the number of pieces of map-generating data to be received in a second period ahead after the first period, based on history of traffic volume under each environmental condition or history of the number of pieces of map-generating data previously received for the road section.