Intelligent cleaning robot
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current home cleaning robots are unable to effectively detect obstacles on the ground for navigation and mapping, leading to inefficient coverage of areas, with some regions being missed or covered multiple times.
Innovation Solution
The intelligent cleaning robot employs stereo cameras with a static projected light pattern for 3D mapping and obstacle avoidance, combined with optical sensors, laser light striping, and time-of-flight laser distance sensors for comprehensive environment sensing and navigation, allowing for autonomous mapping and cleaning of multiple rooms.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If bump sensors and proximity sensors are used for obstacle avoidance, then the robot can detect obstacles, but it cannot detect obstacles on the ground without engaging with them, leading to limited navigational capabilities
Solution Approach 1:
The patent transitions from 2D surface sensors (bump and proximity sensors) to 3D spatial sensing using stereo cameras and structured light. This dimensional change enables the robot to detect obstacles on the ground plane without physical contact, providing depth information and spatial awareness that prevents collisions while maintaining navigation efficiency.
Solution Approach 2:
The patent replaces mechanical bump sensors with optical sensing systems (stereo cameras, structured light projectors). This substitution eliminates the need for physical engagement with obstacles, allowing the robot to detect and avoid obstacles on the ground plane through visual and optical methods rather than mechanical contact.
2Productivity
If a single camera or optical sensor is used for mapping and localization, then the robot can perform visual SLAM, but it cannot effectively detect obstacles on the ground, resulting in incomplete coverage of areas
Solution Approach 1:
The patent merges multiple sensing modalities (stereo vision, structured light, time-of-flight sensors, optical sensors) into a unified sensing system. This combination enables simultaneous achievement of accurate obstacle detection on the ground plane and effective mapping/localization, resolving the contradiction between productivity and measurement precision by integrating complementary sensing capabilities.
Solution Approach 2:
The patent enhances single-camera 2D imaging with stereo vision and structured light to create 3D spatial understanding. This dimensional enhancement allows the system to detect obstacles on the ground plane while maintaining mapping and localization capabilities, achieving both high productivity and accurate obstacle detection simultaneously.
3Productivity
If the robot traverses regions without effective obstacle detection, then it can cover areas, but some areas will be missed or covered multiple times, reducing cleaning efficiency
Solution Approach 1:
The patent implements continuous feedback loops where stereo cameras and structured light systems constantly monitor the environment, providing real-time spatial awareness and obstacle detection. This feedback enables the robot to adjust its traversal path dynamically, ensuring complete and efficient area coverage without missing or repeatedly cleaning the same areas.
Solution Approach 2:
The patent uses 3D structured light and stereo vision to create accurate spatial maps of the environment. This three-dimensional spatial awareness allows the robot to plan and execute efficient traversal paths that ensure complete area coverage, preventing both missed areas and redundant cleaning by understanding the true spatial layout.
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
The robot achieves efficient and thorough coverage of domestic environments by accurately mapping and avoiding obstacles, ensuring all areas are cleaned without repetition, using a combination of 3D data, visual odometry, and sensor fusion for precise navigation and localization.
Implementation Method 1
The intelligent cleaning robot uses stereo cameras with a static projected light pattern to generate 3D data used for the purposes of mapping, localization, and obstacle avoidance
Implementation Method 2
The system also uses multiple time-of-flight (ToF) laser distance sensors oriented in various directions for cliff detection, wall following, obstacle detection, and general perception
Implementation Method 3
Another embodiment of the robot uses a downward facing camera housed within the robot that tracks features on the ground, providing visual odometry from which a motion estimate that is independent of wheel slip may be derived
Data Source
AI summary
An intelligent, autonomous interior cleaning robot capable of autonomously mapping and cleaning multiple rooms in a house in an intelligent manner is described. Various combinations of passive and active sensors may be used to perform mapping, localization, and obstacle avoidance. In particular, the robot uses stereo cameras with a static projected light pattern to generate 3D data. In addition, the robot may use optical sensors in various locations, laser ToF sensors, inertial measurement units and visual odometry to enhance the localization and mapping capabilities.