Geofence Grid Identification via Partition Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems face difficulties in efficiently identifying grids in large geographical regions, particularly when the geofence crosses multiple regional partitions, due to differing grid indexes and the complexity of mapping scales, which hinders services like autonomous driving and online-to-offline services.
Innovation Solution
An artificial intelligence system that loads a map into a cache, segments enclosed areas crossing multiple partitions into sub-areas within single partitions, determines boundary and internal grids, and collects these to identify grids within the geographical region, using processor instructions and geofence information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the map is divided into multiple independently indexed partitions, then the grid identification can be more manageable within each partition, but the complexity increases when the geofence crosses multiple partitions
Solution Approach 1:
The patent segments the geofence identification process by dividing the geofence into multiple sub-geofences, where each sub-geofence is contained within a single partition. This allows each sub-geofence to be independently processed using simple grid indexing, avoiding the complexity of handling跨-partition queries while maintaining overall system efficiency.
2Area of stationary object
If the geofence covers a large geographical area, then more comprehensive service coverage is achieved, but the number of grids to search becomes astronomical
Solution Approach 1:
The patent divides a large geofence into multiple smaller sub-geofences based on partition boundaries. Each sub-geofence contains a manageable number of grids within a single partition, reducing the search complexity from astronomical to linear while maintaining complete coverage of the original large geofence area.
3Adaptability or versatility
If grid indexes are independently assigned in each partition, then each partition can be processed independently, but grids crossing partition boundaries become difficult to identify
Solution Approach 1:
The patent resolves the boundary identification problem by segmenting the geofence into sub-geofences that are each contained within single partitions. This approach leverages the independence of partition indexing while avoiding boundary crossing issues, as each sub-geofence can be processed using its parent partition's independent index system without needing to handle cross-partition grid identification.
Data Source
AI summary
The present disclosure includes artificial intelligence systems and methods for identifying grids for a geofence in an area in a map that includes a plurality of independently indexed regions of grids. The map is gridded into a plurality of grids and divided into a plurality of partitions. The methods include obtaining information of the geofence in the map, and determining a boundary grid series of the geofence from the plurality of grids. The methods include identifying at least one enclosed area in the geographical region, and upon determining that an enclosed area crosses two or more partitions, segmenting the enclosed area into two or more sub-areas. The methods also include determining boundary grids for each sub-area, and identifying internal grids in the sub-area for each sub-area. The methods further include identifying grids in the geographical region by collecting the boundary grids and the internal grids.