Grid Map-Based Doorsill Detection for Autonomous Mobile Devices
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current autonomous mobile devices, such as cleaning robots, face difficulties in detecting doorsills due to varying heights and shapes, leading to potential sticking or missed cleaning areas, and existing solutions require high hardware and software capabilities, increasing costs and risking false detections.
Innovation Solution
A method utilizing a grid map-based environmental mapping, where the autonomous device performs limited movement to detect doorsills using LiDAR, structured light, or visual sensors, clustering grids to identify doorsills based on obstacle information and sensor data, without significantly increasing hardware requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If camera-based object recognition is used to detect doorsills, then doorsill location can be determined, but computing power requirements increase and cost increases
Solution Approach 1:
The patent replaces camera-based visual recognition with a LiDAR-based spatial mapping system. Instead of using complex image processing algorithms to recognize door profiles, the system uses LiDAR to directly measure spatial distances and construct 3D maps, substituting optical-mechanical recognition with direct spatial measurement.
Solution Approach 2:
The patent introduces an environmental map as an intermediary data structure between the LiDAR sensor and the doorsill detection logic. The map stores spatial information about the environment, allowing the system to infer doorsill locations from pre-built environmental data rather than performing real-time complex analysis of raw sensor data.
2Reliability
If deep learning methods are used for obstacle recognition, then recognition capability is improved, but recognition rate is limited and false detections occur
Solution Approach 1:
The patent performs preliminary environmental mapping before the actual cleaning task. The LiDAR sensor builds a complete environmental map of the space beforehand, storing spatial information about walls, obstacles, and potential doorsills. This pre-processing allows the system to have accurate spatial reference data available during operation, eliminating the need for real-time deep learning analysis.
Solution Approach 2:
The system uses the pre-built environmental map as feedback to guide real-time navigation and obstacle detection. By comparing current LiDAR measurements against the stored environmental map, the system can identify changes or anomalies that indicate doorsills, providing a feedback mechanism that reduces false detections compared to standalone deep learning approaches.
3Measurement precision
If structured light sensor is used for 3D mapping, then doorsill detection capability is improved, but hardware cost and complexity increase
Solution Approach 1:
The patent substitutes structured light sensors with LiDAR technology. Both technologies perform 3D mapping, but LiDAR uses laser time-of-flight measurement which is more mature, cost-effective, and energy-efficient for mobile robotic applications compared to structured light projection and camera-based reconstruction.
Solution Approach 2:
The patent changes the measurement parameter from optical pattern recognition (structured light) to direct time-of-flight distance measurement (LiDAR). This parameter change simplifies the hardware requirements while maintaining the ability to detect spatial features like doorsills, as LiDAR directly provides distance information without requiring complex light pattern analysis.
4Reliability
If hardware performance is increased to improve obstacle clearance capability, then negotiation capability is improved, but device cost increases
Solution Approach 1:
The patent performs preliminary environmental mapping and doorsill detection before the cleaning robot attempts to navigate the area. By identifying doorsills in advance using LiDAR and the environmental map, the system can plan appropriate navigation strategies, reducing the need for high-performance mechanical components during actual obstacle negotiation.
Solution Approach 2:
The system uses its own LiDAR sensor and environmental map data to detect and identify doorsills, rather than requiring external assistance or more complex mechanical sensing components. The existing LiDAR infrastructure is leveraged for dual purposes: both navigation and doorsill detection, making the system self-sufficient without additional expensive hardware.
Data Source
AI summary
The present disclosure provides a detection method for an autonomous mobile device, an autonomous mobile device, and a storage medium. The detection method for the autonomous mobile device includes: obtaining an environmental map of an environment in which the autonomous mobile device is located, the environmental map being a grid map; obtaining grids-to-be-processed in the environmental map; clustering the grids-to-be-processed to obtain one or more groups of grids-to-be-processed; for each group of grids-to-be-processed, determining whether the group of grids-to-be-processed corresponds to a doorsill based on environmental information.