Agricultural Landmark Navigation Without Pre-Mapped Fields
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing agricultural navigation systems struggle with autonomous operation in fields without pre-generated maps, particularly in environments with inconsistent or irregular features, leading to difficulties in identifying landmarks for navigation.
Innovation Solution
A landmarking navigation system that scans, detects, and identifies landmarks in real-time using GNSS-based navigation, transitioning to landmark-based navigation when GNSS signals are lost, and maintains operation through inertial measurement units and intermediate navigation methods.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If pre-generated maps with landmarks are created using drones or UAVs before autonomous operation, then navigation accuracy is improved, but device complexity and time consumption increase due to additional vehicles and laborious mapping steps
Solution Approach 1:
The agricultural vehicle performs its own landmark detection and mapping functions using its onboard sensors during normal operation, eliminating the need for separate drone-based mapping operations. The vehicle serves itself by detecting landmarks and creating or updating maps autonomously while conducting agricultural tasks.
Solution Approach 2:
The agricultural vehicle's sensor system serves multiple functions: it detects landmarks for navigation, monitors the agricultural environment, and contributes to map creation or updating. This multi-functionality eliminates the need for dedicated mapping equipment like drones, reducing overall system complexity.
2Measurement precision
If pre-generated maps with landmarks are created using drones or UAVs before autonomous operation, then navigation accuracy is improved, but loss of time increases due to laborious mapping steps
Solution Approach 1:
The system performs preliminary landmark detection and map updates during previous operational passes through the field. By continuously detecting and cataloging landmarks during normal agricultural operations, the map is progressively built and refined without requiring dedicated pre-mapping time.
Solution Approach 2:
Landmark detection and map creation occur continuously during agricultural operations rather than as separate pre-requisites. The vehicle detects landmarks and updates maps throughout its operational life, converting what would be discrete time-consuming mapping operations into continuous background processes.
3Measurement precision
If autonomous operation relies on consistent features like grapevines, then landmark identification accuracy is improved, but adaptability worsens in fields with barren soil, tilled soil, or vegetative stage crops that lack consistent features
Solution Approach 1:
The system adapts its landmark detection parameters based on environmental conditions. In fields with sparse or variable features, the system adjusts detection sensitivity, cataloging a broader range of potential landmarks including less consistent features, while maintaining stricter verification criteria to preserve identification accuracy across diverse agricultural environments.
4Quantity of substance
If non-crop based features like fences, rocks, or trees are used as landmarks, then landmark availability is improved in fields without crops, but identification reliability worsens due to varied profiles, colors, shapes, and distances
Solution Approach 1:
The system segments the landmark identification process into multiple stages: initial detection of potential landmarks, classification by feature type, verification against catalogued characteristics, and confirmation for navigation use. This segmented approach allows the system to handle diverse landmark types while maintaining reliability through progressive filtering and verification.
Solution Approach 2:
The system uses feedback from successful landmark identifications to refine its detection algorithms and update the landmark catalog. By continuously learning from confirmed landmarks and adjusting detection parameters based on verification results, the system improves its ability to reliably identify varied non-crop features across different environmental conditions.
Data Source
Figure 1A
Figure 1B
Figure 1C
AI summary
A landmarking navigation system includes one or more sensors that observe one or more features in a field or proximate to a field. One or more processors are in communication with the one or more sensors. The one or more processors landmark the one or more features. Landmarking includes identifying the one or more features as landmarks, respectively. The landmarks are catalogued as catalogued landmarks. The one or more processors landmark navigate the agricultural vehicle. Landmark navigating includes observing the one or more features in the field with the sensors and comparing the observations with the catalogued landmarks. The features are identified as the catalogued landmarks, respectively. Vehicle kinematics (one or more of position, heading, speed, pitch, yaw or roll) of the agricultural vehicle are determined relative to the identified catalogued landmarks.