Vision-Based Lawn Mower Navigation for Temporary Obstacle Avoidance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing autonomous lawn mowers rely on boundary wires for navigation, which cannot address temporary obstacles and periodic changes in a lawn, leading to potential damage from collision sensors that struggle with irregularities like dead grass patches or holes.
Innovation Solution
A vision-based navigation system using cameras and processors to analyze images in real-time, categorizing image sections as mowable or unmowable, and adjusting the lawn mower's course to avoid obstacles, allowing for continuous operation without the need for boundary wire revisions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a boundary wire is used to define lawn bounds and identify permanent obstacles, then the autonomous lawn mower can navigate within bounds and avoid permanent obstacles, but the system cannot address temporary obstacles and periodic changes in the lawn
Solution Approach 1:
The lawn area is divided into multiple image sections, with each section independently analyzed for mowable versus unmowable content. This segmentation allows the system to identify temporary obstacles in specific regions without being constrained by a fixed boundary wire layout.
Solution Approach 2:
The mechanical boundary wire system is replaced with a vision-based navigation system using cameras and image processing. This substitution enables the system to detect both permanent and temporary obstacles through optical sensing rather than relying on physical wire boundaries.
2Adaptability or versatility
If a collision sensor is used to negotiate temporary obstacles, then the system can detect some irregularities, but over time this results in damage to the encountered obstacles or the lawn mower itself
Solution Approach 1:
The vision system performs preliminary detection of obstacles before the lawn mower reaches them. By identifying temporary obstacles such as dead grass patches, holes, and seasonal ornaments in advance, the system can plan avoidance maneuvers that prevent collision and damage.
Solution Approach 2:
The system continuously captures images and processes them in real-time to provide feedback on the lawn conditions ahead. This feedback loop enables dynamic adjustment of the mower's path to avoid temporary obstacles while maintaining efficient mowing operations.
3Adaptability or versatility
If the collision sensor is the primary method for detecting obstacles, then the system can operate without boundary wire revisions, but the collision sensor cannot sense many periodic irregularities like dead grass patches or holes
Solution Approach 1:
The system changes the detection parameter from mechanical contact (collision sensor) to optical sensing (camera). This parameter change enables the detection of various obstacle types including dead grass patches, holes, and seasonal ornaments that are invisible to collision sensors until contact occurs.
Data Source
AI summary
Autonomous vehicle navigation is disclosed. An example lawn mower includes a vision sensor configured to capture an image of an area being approached. The lawn mower includes a processor in communication with the vision sensor and configured to divide the image into a grid of separate image sections. Each of the image sections corresponds to a respective portion of the area being approached. For each of the image sections in the grid the processor is further configured to calculate a percentage of the image section categorized as corresponding with unmowable material and assign a result value based on a comparison between the calculated percentage and one or more movement thresholds. The lawn mower includes a drive system in communication with the processor and configured to control movement of the lawn mower based on an arrangement of the result values in the grid of the image sections.