Unequal-Interval Map Grids for Lower-Size Robot Navigation Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The increasing data size of map data for autonomous work vehicles poses challenges in handling and hardware requirements, with conventional methods requiring quadrupled parcels when doubling the number of rows and columns, leading to inefficient data storage and increased hardware needs.
Innovation Solution
Representing map data using a grid of parcels with assigned data sets that include characteristic and grid values, allowing for unequal intervals and integrating parcels with equal characteristic values, thereby reducing the number of parcels needed to represent larger areas.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If map data is represented using a conventional grid with equal intervals, then the map structure is simple and easy to process, but the data size increases significantly when the work area expands
Solution Approach 1:
The map data is segmented into multiple layers, where the first layer contains coarse grid data for the entire work area and the second layer contains detailed parcel data only for specific regions of interest. This segmentation allows the system to maintain simple processing for general navigation while storing detailed information only where needed, thus reducing overall data size without compromising functionality
Solution Approach 2:
The patent applies local quality by providing different levels of data detail in different regions. The first layer provides uniform coarse-grained data across the entire map, while the second layer provides fine-grained parcel-level data only for specific parcels that require detailed information. This allows the system to optimize data storage by providing detailed information locally where needed while maintaining simplicity in other areas
2Measurement precision
If the number of rows and columns in the grid is doubled to increase map resolution, then the map detail improves, but the number of parcels quadruples leading to increased data storage requirements
Solution Approach 1:
The system dynamically adjusts the level of detail provided for different parcels based on their importance and the current operational context. The second layer of data structures allows the system to include detailed parcel information only for parcels that are currently relevant to the autonomous vehicle's operations, while using coarser representations for less important areas, thus maintaining high effective resolution without quadrupling overall data storage
Solution Approach 2:
The patent merges two different data representation approaches: a regular grid system for general spatial reference and an irregular parcel-based system for detailed regional information. By combining these approaches, the system achieves high measurement precision where needed while avoiding the exponential increase in data storage that would result from uniformly increasing grid resolution across the entire map
3Measurement precision
If detailed map data is stored for the entire work area, then navigation accuracy improves, but hardware specifications must be increased to handle the large data size
Solution Approach 1:
The map data is segmented into two layers: a first layer with coarse grid data that provides basic spatial structure for the entire work area, and a second layer with detailed parcel data that provides high-precision navigation information only where needed. This segmentation allows the autonomous vehicle to achieve high navigation accuracy in relevant areas while keeping overall data size manageable, thus avoiding the need for increased hardware specifications
Solution Approach 2:
The system changes the parameter of data detail level dynamically based on spatial location and operational context. By using characteristic values to identify important parcels and providing detailed data only for those parcels in the second layer, the system maintains high navigation accuracy parameters where needed while reducing overall data volume, thereby avoiding the need for upgraded hardware
Data Source
AI summary
A map is stored in a robot that has a travelling unit, wherein the map is represented by a grid having a plurality of parcels, and a data set is assigned to each parcel such that the grid is represented by a plurality of data sets, each data set assigned to each parcel includes a characteristic value representing a characteristic of the parcel, and a grid value representing a shape of the parcel.