UAV Docking With Visual Fiducials for GPS-Denied Landing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing unmanned aerial vehicle (UAV) base stations are large, mechanically complex, and expensive, requiring accurate GPS positioning and complex mechanical methods for docking, which limits their functionality in GPS-denied environments and necessitates frequent human intervention for battery swapping.
Innovation Solution
A small, passive funnel-shaped nest with spring contacts and visual tags, supplemented by larger fiducials, enables precise UAV landings and charging without complex actuation, using visual tracking and control software to maintain accurate positioning and communication in various conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If large base station enclosures are used, then UAV docking reliability is improved, but device complexity and cost increase
Solution Approach 1:
The patent replaces complex mechanical docking systems with vision-based guidance. Instead of using large enclosures and mechanical alignment mechanisms, the system uses visual fiducial markers and computer vision algorithms to guide the UAV to the docking station, eliminating the need for complex mechanical structures while maintaining docking reliability
Solution Approach 2:
The patent uses visual fiducial markers (copies of known patterns) on the docking station that the UAV's vision system can recognize and track. These markers serve as optical copies of the docking target, allowing the UAV to locate and align with the docking station without requiring large physical enclosures or complex mechanical guidance systems
2Measurement precision
If accurate GPS positioning is required, then landing precision is improved, but adaptability to GPS-denied environments deteriorates
Solution Approach 1:
The patent introduces visual fiducial markers as an intermediary between the UAV and the docking station. These markers serve as a mediator that the UAV's vision system can detect and use for precise localization, replacing GPS as the positioning reference and enabling accurate docking in GPS-denied environments
Solution Approach 2:
The patent substitutes GPS-based positioning with vision-based positioning using fiducial markers. This replacement allows the system to achieve high positioning precision without relying on GPS satellite signals, thereby enabling operation in GPS-denied environments such as indoors or in urban canyons
3Manufacturing precision
If complex mechanical methods are used for docking, then docking precision is improved, but device complexity and cost increase
Solution Approach 1:
The patent replaces mechanical docking guidance systems with a vision-based approach. Instead of using mechanical guides, rails, or articulated arms for alignment, the system uses computer vision to detect fiducial markers and calculate the UAV's position and orientation, achieving precise docking without complex mechanical structures
Solution Approach 2:
The system enables the UAV to autonomously locate and align with the docking station using its own onboard cameras and vision processing algorithms. The UAV independently detects fiducial markers, computes its pose, and navigates to the docking position without requiring external mechanical guidance or manual intervention
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
Facilitates reliable, autonomous UAV operations with reduced human intervention, enabling efficient battery management and extended operation time, while being cost-effective and functional in GPS-denied environments.
Implementation Method 1
detecting the fiducial in at least one of the one or more images
Implementation Method 2
control the propulsion mechanism to cause the unmanned aerial vehicle to fly to a first location
Implementation Method 3
control the propulsion mechanism to cause the unmanned aerial vehicle to land on the landing surface
Data Source
Figure 1
Figure 2
Figure 3A
AI summary
Described herein are systems for automated docking of an unmanned aerial vehicle. For example, some systems include an unmanned aerial vehicle including a propulsion mechanism, an image sensor, and processing apparatus; and a dock including a landing surface configured to hold the unmanned aerial vehicle and a fiducial on the landing surface, wherein the processing apparatus is configured to: control the propulsion mechanism to cause the unmanned aerial vehicle to fly to a first location in a vicinity of the dock; access one or more images captured using the image sensor; detect the fiducial in at least one of the one or more images; determine a pose of the fiducial based on the one or more images; and control, based on the pose of the fiducial, the propulsion mechanism to cause the unmanned aerial vehicle to land on the landing surface.