Multi-floor Map Alignment via Anchor Points
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current SLAM systems are inadequate for mapping large, multi-floor industrial buildings, as they lack the accuracy and precision required for such complex environments, having been primarily developed for limited areas in academic research.
Innovation Solution
A method that involves building multiple maps of different floors by detecting anchor points with the same (x, y) coordinates, linking these points to align the maps, and optimizing the alignment using error correction, allowing for the creation of a comprehensive map of a multiple-floor structure using a single robot.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If academic SLAM systems are used for limited areas, then implementation simplicity is maintained, but mapping accuracy and precision for large multi-floor buildings is insufficient
Solution Approach 1:
The patent divides the large multi-floor building into multiple smaller maps, each representing a specific floor or area. These individual maps are then aligned and integrated using anchor points to form a comprehensive building-wide map. This segmentation approach allows the system to manage complexity while maintaining high mapping precision across the entire structure.
Solution Approach 2:
The patent introduces anchor points as intermediary elements that serve as reference markers across different floors and areas. These anchor points facilitate the alignment and integration of multiple individual maps, enabling accurate multi-floor mapping without requiring direct complex interactions between all map elements.
2Area of stationary object
If multiple maps are created for different floors, then comprehensive coverage is achieved, but alignment precision between maps becomes challenging
Solution Approach 1:
The patent establishes a common reference frame for all floor maps by using anchor points with consistent (x, y) coordinates across different floors. This creates an equipotential alignment basis where each map can be precisely positioned relative to others, ensuring consistent spatial relationships throughout the multi-floor structure.
Solution Approach 2:
The patent implements error correction mechanisms that use the anchor points to detect and correct alignment errors between maps. By continuously referencing the anchor points during map integration, the system can identify and rectify positioning discrepancies, maintaining high alignment precision across all floors.
3Device complexity
If a single robot is used for mapping, then system cost is reduced, but mapping time and productivity decrease
Solution Approach 1:
The patent employs a single robot that performs mapping operations in multiple passes, systematically covering different floors and areas. The robot collects data incrementally across multiple visits, allowing comprehensive mapping without requiring multiple simultaneous robots, thus reducing system complexity while maintaining productivity.
Solution Approach 2:
The patent enables continuous mapping operations by having the robot repeatedly visit and map different areas of the building. The system accumulates map data over time through continuous operation, allowing a single robot to achieve complete building coverage through persistent, ongoing mapping activities rather than requiring multiple robots operating simultaneously.
Data Source
AI summary
A method for mapping a multiple-floor structure includes building a plurality of maps that includes one map for each of at least two of the floors, detecting a set of anchor points, where the set of anchor points includes at least one anchor point on each of the maps, linking the anchor points to produce a set of linked anchor points, and aligning the maps around the set of linked anchor points to produce a set of aligned maps. A method for mapping a structure using a single robot includes generating, by the robot, a first map at a first point in time, storing the first map, generating, by the same robot, a second map at a second point in time subsequent to the first point in time, and aggregating the first map and the second map to produce an aggregated map.