Robot Environment Map Merging Using Shared Anchor Points
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for creating a combined environmental map for self-propelled soil cultivation devices face difficulties when the overlapping area of partial maps is small, with insufficient characteristic features.
Innovation Solution
Self-propelled tillage implements identify each other using unique identifiers, create local environment maps, and merge these maps into a global map by aligning them based on shared anchor points, allowing for a common coordinate system without requiring extensive overlap.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional environmental map merging methods are used based on characteristic features, then map integration is possible, but the process fails or becomes unreliable when the overlapping area is small with insufficient characteristic features
Solution Approach 1:
The patent introduces artificial anchor points with unique identifiers as intermediary objects to facilitate map merging. These anchor points serve as reliable reference markers that enable accurate coordinate system alignment between multiple environment maps even when the overlapping area is small, thus resolving the contradiction between merging reliability and overlapping area size
Solution Approach 2:
The patent changes the approach from relying on environmental characteristic features to using artificial anchor points with unique identifiers. This parameter change transforms the basis of map alignment from natural features (which require large overlap areas) to controlled artificial markers (which work with minimal overlap), thereby improving reliability independent of overlapping area size
2Adaptability or versatility
If multiple soil cultivation devices create and merge environment maps independently, then collaborative work is enabled, but the complexity of coordinate system alignment and map integration increases
Solution Approach 1:
The patent uses anchor points with unique identifiers as intermediary reference objects that simplify the coordinate system alignment process. Each device detects anchor points and uses them to transform coordinates to a common reference system, greatly reducing the integration complexity compared to traditional feature-based methods while enabling collaborative work among multiple devices
Solution Approach 2:
The patent employs a centralized server that stores and manages environment maps from multiple devices. The server creates and maintains a unified global environment map by integrating individual device maps, copying and consolidating data centrally rather than requiring complex peer-to-peer coordination, thus reducing overall system complexity while maintaining collaborative capability
Data Source
Figure 1
Figure 2
AI summary
The invention relates to a method for creating an environment map (3, 4, 5) for a soil cultivation device (1, 2), wherein a first soil cultivation device (1) detects first feature data of an environment and processes it into a first environment map (3), wherein a second soil cultivation device (2) detects second feature data of the environment and processes it into a second environment map (4), and wherein the first environment map (3) and the second environment map (4) are combined to form a common environment map (5). In order to merge the environment maps of several robots even with only a small overlap area, it is proposed that the first soil cultivation device (1) identifies an object (6) as an anchor point of the environment known to the first soil cultivation device (1) based on a unique identifier (7) of the object (6) and stores relative position information (9) of the anchor point in a coordinate system of the first environment map (3).wherein the second tillage implement (2) recognizes the same object (6) as an anchor point known to the second tillage implement (2) and stores relative position information (9) of the anchor point in a coordinate system of the second environment map (4), and wherein the first environment map (3) and the second environment map (4) are combined into a common global environment map (5) based on the relative position information (9) of the anchor point contained therein, and wherein the unique identifier (7) is a code uniquely identifying the object (6), namely an optical code, an electronic code, a magnetic code, a shape code and/or a color code.