Field Robot Navigation Using Traversability Mapping Offline
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional autonomous navigation systems for agricultural robots face challenges in real-world scenarios due to uneven terrains and low network connectivity, which can lead to inaccurate navigation and potential damage from obstacles like muddy or dusty conditions.
Innovation Solution
The system employs a method that uses GPS, GNSS, machine learning models, and a Kinodynamic Motion Planning Model (KMPM) to determine coefficients of traversal for field robots, predicting their motility based on terrain and obstacles, and identifies traversable routes, while also utilizing transfer learning to adapt models from path planning robots for improved accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If GPS and cloud-based image processing are used for autonomous navigation, then navigation capability is provided, but accuracy and reliability deteriorate due to low network connectivity in remote agricultural areas
Solution Approach 1:
The patent introduces an on-board processor as an intermediary that locally processes images captured by cameras mounted on the robot. This intermediary enables the robot to perform autonomous navigation by processing visual data and determining its position and orientation without relying on cloud-based systems, thus maintaining navigation reliability in remote areas with low network connectivity
Solution Approach 2:
The patent replaces the cloud-based image processing system with an on-board processing system. Instead of transmitting images to remote servers for processing, the robot uses its own processor to analyze images and determine navigation parameters, substituting a mechanical/local system for a remote/cloud-based system to ensure reliable operation in offline conditions
2Ease of operation
If traditional navigation systems are used, then basic navigation is achieved, but accuracy deteriorates on uneven and distinct terrains in agricultural fields
Solution Approach 1:
The patent applies local quality by using multiple cameras positioned at different locations on the robot to capture images of distinct terrain features. By processing these localized views and identifying unique terrain characteristics in each camera's field of view, the system achieves high location accuracy on uneven and distinct agricultural terrains
Solution Approach 2:
The patent transitions from traditional 2D GPS-based navigation to a multi-dimensional approach by capturing images from multiple camera angles and processing them to extract 3D terrain information. This dimensional enhancement allows the robot to accurately locate itself on complex terrains by matching multi-view image features with preprocessed terrain maps
3Measurement precision
If more sophisticated navigation systems are deployed, then navigation accuracy improves, but device complexity increases
Solution Approach 1:
The patent achieves high location accuracy with moderate complexity by making the image processing system multi-functional. The same on-board processor and camera system used for navigation also enable the robot to identify terrain types, detect obstacles, and adapt its motion planning, eliminating the need for separate specialized systems for each function
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
This approach enables more accurate and efficient autonomous navigation of field robots, reducing the risk of collision and improving task completion in diverse agricultural environments, even under challenging conditions.
Implementation Method 1
The direction may be determined using a Global Positioning System (GPS), and a Global Navigation Satellite System (GNSS) based on the current location of the FR, and the location of a checkpoint or the target location
Data Source
AI summary
A system and a method for autonomous navigation of a field robot (FR) is disclosed. The system receives a current location of the FR, a target location, and a sequence of checkpoints within a field. The system then determines a direction of navigation for the FR based on the current location, the target location, and the sequence of checkpoints. Further, the system obtains a set of images of parts of the field from cameras installed on the FR. Subsequently, the system determines a coefficient of traversal of the parts of the field in the set of images. The system determines one or more traversable areas based on the coefficient of traversal. Finally, the system identifies a traversable route for the FR to navigate to a checkpoint or the target location.