Vision-Guided Windrow Clearing for GPS-Denied Autonomous Machines
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous machines face challenges in steering and clearing windrows, particularly in environments unsuitable for human operators, due to complex steering and traction issues, and existing systems relying on high-cost mmWave sensors and GPS, which can provide unreliable results in urban or confined areas.
Innovation Solution
A method and system utilizing a visual perception sensor, such as a stereo camera or LiDAR, mounted on the machine to detect windrows and generate control commands for steering, propulsion, and implement control, allowing autonomous or semi-autonomous operation without the need for expensive GPS systems.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If GPS-based autonomous navigation is used, then autonomous operation capability is improved, but system cost and reliability in urban/tunnel environments worsen
Solution Approach 1:
The patent introduces visual perception sensors (cameras, LiDAR) as intermediary devices that detect windrow features and generate navigation guidance. These sensors serve as mediators between the machine's autonomous navigation system and the physical windrow obstacles, enabling reliable operation in GPS-denied environments by providing alternative visual cues for path planning and steering control.
Solution Approach 2:
The patent replaces GPS-based mechanical/electronic navigation systems with vision-based detection and control. By substituting the GPS dependency with onboard visual sensors that process real-time images of windrows, the system achieves autonomous navigation in environments where GPS signals are unavailable or unreliable, thereby improving both reliability and reducing system cost.
2Measurement precision
If mmWave sensors and GPS systems are used, then windrow detection capability is improved, but system cost increases
Solution Approach 1:
The patent replaces expensive mmWave sensors and GPS equipment with relatively inexpensive visual perception sensors (standard cameras, LiDAR). These visual sensors provide sufficient measurement precision for windrow detection and navigation while significantly reducing system cost. The vision-based approach uses readily available, low-cost hardware that can achieve the required detection accuracy without the high expenses associated with specialized radar or GPS systems.
Solution Approach 2:
The patent uses visual perception to create a digital representation (copy) of the windrow structure and position through image processing. By capturing and analyzing visual features of windrows through cameras or LiDAR, the system generates sufficient navigation data without requiring expensive specialized sensors, thereby achieving accurate windrow detection at lower system cost.
3Adaptability or versatility
If articulated steering with blade operation is implemented, then task versatility is improved, but steering control complexity increases
Solution Approach 1:
The patent implements a feedback control system where visual perception sensors continuously monitor windrow position and machine orientation. The controller processes this visual feedback information and dynamically adjusts steering commands for the articulated joints and blade position. This closed-loop feedback mechanism simplifies the control of complex articulated steering by using real-time visual information to automatically coordinate multiple control parameters, thereby managing steering control complexity while maintaining task versatility.
Solution Approach 2:
The patent employs a unified visual perception-based control system that simultaneously manages articulated steering, blade operation, and navigation. The single vision-based control architecture performs multiple functions (steering control, blade positioning, path planning) that would otherwise require separate specialized systems, thereby reducing overall control complexity while maintaining the versatility needed for complex tasks involving articulated mechanisms and blade operations.
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
Enables efficient and precise windrow clearing by machines like motor graders, maintaining productivity in various conditions without human operators, reducing costs by using onboard sensors and existing hardware, and overcoming GPS limitations in challenging environments.
Implementation Method 1
detecting the windrow via a visual perception sensor coupled to a controller
Data Source
AI summary
A controller-implemented method for automated control of at least one of an autonomous or semiautonomous machine to clear at least one windrow along a work surface. The method includes utilizing a visual perception sensor to detect said windrow, generating a plurality of windrow visual detection signals, determining a position of a windrow relative to the machine based upon the plurality of windrow visual detection signals, and generating machine control commands to clear the windrow.