Cleaning Robot Multi-Floorplan Recognition Using LIDAR and SLAM
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current cleaning robots require users to delete existing maps and no-go lines when switching between floors or environments, leading to inefficiencies in mapping and cleaning operations.
Innovation Solution
A cleaning robot that automatically determines its floorplan by localizing and matching its environment to stored maps, using sensors like accelerometers and gyroscopic sensors to detect motion and LIDAR for environment scanning, and implementing a two-step process of localization and place recognition to select the best match with a confidence threshold, allowing for multiple persistent maps with independent no-go lines.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If the robot stores multiple floorplans and automatically determines current location, then adaptability to different floors is improved, but device complexity increases
Solution Approach 1:
The system segments the floor recognition problem into two independent modules: localization (determining robot position using LIDAR and SLAM) and place recognition (matching detected environment to stored floorplans). This segmentation allows each module to be optimized independently, reducing overall system complexity while maintaining multi-floor adaptability.
Solution Approach 2:
The robot performs preliminary localization before attempting place recognition. By first determining its position through LIDAR scanning and SLAM algorithms, the robot narrows down the search space for floorplan matching, making the overall process more efficient and manageable despite storing multiple floorplans.
2Extent of automation
If the robot performs localization and place recognition to determine current floor, then automatic floor identification is improved, but loss of time increases
Solution Approach 1:
The robot performs a two-step process where it first performs partial localization to get a rough position estimate, then performs place recognition only in the relevant area. This partial action approach avoids the need to compare the entire environment against all stored floorplans, reducing computation time while maintaining automatic identification.
Solution Approach 2:
The system uses feedback from the localization step to guide the place recognition process. The localization results provide initial constraints that feed into the place recognition algorithm, allowing the system to quickly converge on the correct floorplan match without exhaustive searching.
3Measurement precision
If the robot uses LIDAR with limited range for environment detection, then measurement precision is improved, but difficulty of detecting and measuring increases
Solution Approach 1:
The robot compensates for LIDAR's limited range by operating in the time dimension - performing multiple sequential scans as it moves through the environment. By accumulating data over time and multiple positions, the robot builds a complete floorplan representation despite the sensor's restricted instantaneous range, maintaining precision while overcoming detection difficulties.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables seamless switching between floorplans without user intervention, efficient mapping, and flexible cleaning operations by automatically identifying and adapting to new or changed environments, while maintaining user-defined boundaries.
Implementation Method 1
The robot creates the map by starting from a base/charging station, and mapping all the areas it can get to. It uses a LIDAR (Light Detection and Ranging) laser & sensor system
Implementation Method 2
an accelerometer, tilt sensor, gyroscopic sensor or other device is used to determine that there has been no motion indicating that the cleaning robot has been picked up
Implementation Method 3
The robot can compute angular excursions based on integrated angular rates generated by the gyroscopic or other sensors
Data Source
AI summary
In one embodiment, a cleaning robot stores multiple floorplans. The robot automatically determines which floorplan it is in by localizing and trying to match its detected environment to the stored map. The best fit is the map for the current floor. If there is no match above a confidence threshold, the robot assumes it is a new floor or the floorplan has changed (e.g., furniture has moved), and the robot initiates a discovery mode to map the floorplan and add it to the robot's memory.