Autonomous Mower Vision Docking With Visual Zone Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous lawn mowers face challenges in navigating around temporary obstacles and periodic changes in a lawn, as boundary wires alone are insufficient to address these issues without costly revisions, and collision sensors may lead to damage from unexpected obstacles.
Innovation Solution
The implementation of a vision-based navigation system that uses cameras and image processors to analyze real-time conditions, allowing the lawn mower to avoid obstacles and recognize visual identifiers for charging stations or temporary exclusion zones, thereby overriding obstacle-avoidance responses for safe docking and operation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a boundary wire is used to define lawn bounds and identify permanent obstacles, then the autonomous lawn mower can navigate within the bounds, but temporary obstacles and periodic changes in the lawn cannot be addressed without costly and time-consuming revisions to the boundary wire
Solution Approach 1:
The patent replaces the mechanical boundary wire system with a vision-based navigation system using cameras and image processing. This optical system can dynamically detect and adapt to temporary obstacles and lawn changes without requiring physical reconfiguration of boundary wires, thereby resolving the contradiction between adaptability and ease of manufacture.
Solution Approach 2:
The vision-based system provides dynamic obstacle detection and avoidance capabilities, allowing the lawn mower to adapt in real-time to temporary obstacles and periodic lawn changes. This dynamic approach eliminates the need for static boundary wire revisions while maintaining navigation within bounds.
2Reliability
If a collision or bump sensor is used to deal with unexpected obstacles, then the autonomous lawn mower can detect obstacles, but over time it can result in damage to the encountered obstacles or the autonomous lawn mower itself
Solution Approach 1:
The vision-based navigation system performs preliminary obstacle detection and identification before the lawn mower reaches the obstacle. By detecting and classifying objects in advance using image processing, the system can plan avoidance maneuvers that prevent collision entirely, eliminating the harmful effects associated with last-minute collision detection.
Solution Approach 2:
The camera system acts as an intermediary between the lawn mower and physical obstacles, providing advance notice and detailed information about obstacles before contact occurs. This intermediary detection mechanism enables gentle avoidance maneuvers rather than forceful collision-based detection.
3Adaptability or versatility
If a vision-based navigation system is implemented to address temporary obstacles, then the autonomous lawn mower can navigate around them, but the system complexity increases with cameras and image processors
Solution Approach 1:
The vision-based navigation system serves multiple functions simultaneously: it detects temporary obstacles, identifies permanent obstacles, recognizes charging station visual identifiers, and provides navigation guidance. This multi-functionality justifies the added complexity by consolidating multiple detection and navigation tasks into a single integrated system.
Solution Approach 2:
The image processor aboard the lawn mower performs real-time analysis of captured images to identify obstacles and navigation targets autonomously. The system processes and interprets visual data independently, enabling the lawn mower to self-navigate and self-adjust its path without external intervention, thereby managing complexity through autonomous operation.
4Reliability
If the vision-based navigation system perceives the charging station as an obstacle, then obstacle avoidance is maintained, but docking and recharging cannot occur
Solution Approach 1:
The vision system applies different recognition criteria to different objects in the field of view. When the camera detects the specific visual identifier pattern of the charging station, the system switches from obstacle avoidance mode to docking mode. This localized quality change in object recognition allows the same vision system to serve both obstacle avoidance and charging functions without conflict.
Solution Approach 2:
Instead of treating the charging station as a generic obstacle to be avoided, the system inverts the recognition logic: the specific visual identifier of the charging station triggers a相反 response (docking) rather than the standard obstacle response (avoidance). This inverted recognition principle resolves the contradiction by making the charging station an exception to the obstacle avoidance rule.
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 vision-based navigation system effectively addresses temporary obstacles and periodic changes in a lawn, enhancing the autonomous lawn mower's ability to safely navigate and recharge, while minimizing damage to obstacles and the mower itself.
Implementation Method 1
a camera, for example, that continually intakes images as the lawn mower moves forward
Implementation Method 2
an image processor aboard the lawn mower may be configured to determine that the image data represent an obstacle
Implementation Method 3
A boundary wire emitting an electromagnetic field or pulse may be sensed by the autonomous lawn mower
Data Source
AI summary
A lawn vehicle network includes a charging station having a visual identifier, a lawn vehicle having a battery, a blade system, a drive system whose output effects lawn vehicle forward movement, a processor board connected to both systems, the processor board capable of processing image data and sending commands to both systems, and a vision assembly connected to the processor board and able to transmit image data to the processor board, and the processor board, having received the image data, able to, if the image data represent a first object, maintain the drive system's output at the time of that determination, if the image data represent a second object, change the drive system's output at the time of that determination, and if the image data represent the visual identifier, maintain the drive system's output or send a shutoff command to the vision assembly at the time of that determination.