Mower Low-Profile Object Detection with Buffered ROI Sensing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous mowers and off-highway vehicles face challenges in detecting low-profile objects, such as sprinkler lines, which are difficult to identify due to their small size and partial coverage by grass, leading to potential damage and inefficiencies in operation.
Innovation Solution
The implementation of an autonomy controller that processes sensor data from 2D and 3D sensors to detect objects, slows down the vehicle when an object is suspected, buffers sensor data for confirmation, and adjusts the path accordingly, with the option to notify an external system for further feedback when uncertainty exists.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional sensors are used to detect objects on autonomous mowers, then large objects can be detected, but small low-profile objects covered by grass cannot be reliably detected
Solution Approach 1:
The detection area is divided into multiple regions of interest (ROIs) based on depth information from the 3D sensor. Each ROI corresponds to a specific depth range and is processed separately to identify low-profile objects at different distances from the mower, improving detection precision without overwhelming the system with all sensor data simultaneously
Solution Approach 2:
The system merges data from multiple sensor types (2D camera and 3D depth sensor) to create a comprehensive view of the environment. By combining visual information with depth mapping, the system overcomes the limitations of individual sensors and achieves reliable detection of small, grass-covered objects
2Productivity
If the mower operates at high speed, then productivity is improved, but the risk of collision with undetected low-profile objects increases
Solution Approach 1:
The system performs preliminary detection and classification of objects at a distance using the 3D sensor before the mower reaches them. By identifying potential hazards in advance and processing sensor data continuously, the system maintains high operating speeds while ensuring reliable collision avoidance through proactive rather than reactive measures
Solution Approach 2:
The autonomy controller continuously receives feedback from sensors and adjusts mower operation in real-time. When low-profile objects are detected or suspected, the system provides feedback to slow the mower down for closer inspection, and resumes normal speed when the area is clear, thus maintaining both productivity and safety
3Measurement precision
If the mower slows down frequently to inspect suspected objects, then detection accuracy improves, but productivity decreases
Solution Approach 1:
The system applies partial action by only slowing down when specific conditions are met (suspected low-profile object detection), rather than continuously monitoring at low speed. This selective approach maintains high productivity during normal operation while ensuring accurate inspection only when necessary, avoiding unnecessary speed reductions
4Measurement precision
If multiple sensors and processing steps are added to detect low-profile objects, then detection capability improves, but device complexity increases
Solution Approach 1:
The 3D sensor serves multiple functions: it creates depth maps for general navigation, identifies regions of interest containing potential low-profile objects, and provides distance information for speed control. This multi-functionality reduces the need for additional specialized sensors while maintaining high detection precision
Data Source
AI summary
Low profile object detection can be performed on mowers or other vehicles that may be autonomous. An autonomy controller can be employed on a mower to receive and process sensor data for a detection area to determine whether an object may be present in a region of interest within the detection area. When the autonomy controller determines that an object may be present, it can cause the ground speed of the mower to be slowed and can commence buffering region of interest sensor data over a period of time. The autonomy controller can process the buffered region of interest sensor data to determine whether an object is present in the region of interest, and if so, can alter the path of the mower appropriately.