Collaborative Map Updating for Incomplete Indoor SLAM Areas
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing SLAM technologies generate fragmentary, low-resolution maps that require inefficient remapping and resource usage, and GPS is inaccurate for precise indoor navigation, leading to suboptimal device operation.
Innovation Solution
A central processor coordinates multiple devices to collaboratively build and update maps by selecting devices with superior capabilities to capture high-quality image data, optimizing data retrieval, and integrating it into existing maps, enhancing localization accuracy and reducing redundancy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a single device builds maps independently using SLAM technology, then the device can operate autonomously with its own resources, but the resulting maps are fragmentary, low-resolution, and require inefficient remapping
Solution Approach 1:
The patent merges map data from multiple independent devices to create a comprehensive high-resolution master map. The central processor combines fragmentary maps from various devices, eliminating gaps and improving resolution without requiring individual devices to remap entire areas.
Solution Approach 2:
The patent segments the map-building task across multiple devices, where each device captures data for specific regions or perspectives. This division allows parallel data collection from different locations, which is then integrated into a complete high-resolution map.
2Measurement precision
If devices remap already mapped areas to improve data quality, then map resolution improves, but camera, processing and memory resources are wasted
Solution Approach 1:
The central processor analyzes the quality and completeness of received map data, providing feedback to determine which areas need additional data collection. This feedback mechanism prevents redundant remapping of already sufficient areas while directing resources to areas that actually need improvement.
Solution Approach 2:
Instead of requiring complete remapping of entire areas, the system applies partial action by selectively capturing additional data only in specific regions where resolution or completeness is insufficient, based on central processor evaluation.
3Productivity
If a flying drone captures map data quickly, then sufficient initial map data is obtained ahead of other devices, but fewer frames of objects are captured yielding less data
Solution Approach 1:
The system combines the fast-captured rough map from drones with detailed data from stationary or slower-moving devices. The drone provides initial coverage and identifies gaps, while other devices capture additional frames and detailed information to supplement the quick overview.
Solution Approach 2:
The drone performs preliminary action by capturing an initial rough map quickly, establishing baseline coverage before other devices perform detailed data collection. This preliminary mapping guides subsequent data gathering efforts to focus on areas needing improvement.
4Measurement precision
If GPS is used for device navigation, then devices can locate themselves outdoors, but precision is insufficient for indoor navigation requiring accuracy within a few inches
Solution Approach 1:
The patent introduces map features and visual landmarks as intermediary reference points for localization. Instead of relying solely on GPS satellites, devices use the captured map data and recognizable environmental features as intermediate references to achieve precise positioning indoors where GPS is unavailable.
Data Source
AI summary
A collaborative map building method is described in which if it is determined, for example, by a SLAM network edge, that a portion of a map corresponding to a target area of a first region is incomplete, a request to a device is transmitted to move toward the target area. The map is updated based on the scan performed by the device in the target area. Also described is a virtual image guide embodiment in which upon determining that a map of a target area of the first region is incomplete, a virtual guide image is generated on a device to appear at a target location. The user of the device is guided to the target area by the virtual image.