3D UAV Delivery Path Representation for Obstacle-Aware Access
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Unmanned aerial vehicles (UAVs) face challenges in detecting delivery surfaces and navigating around obstacles during package delivery, as existing technologies struggle to extract and encode safe delivery paths from available 2D and 3D data.
Innovation Solution
A system that determines a three-dimensional variable width delivery path for UAVs, transects it with planar surfaces, and constructs a delivery path data object using ordered lists of shapes to represent the path, incorporating LIDAR data and aerial imagery to avoid obstacles and adhere to regulatory restrictions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If 2D and 3D data is used to represent delivery paths, then delivery path information can be stored, but extracting and encoding safe delivery paths becomes challenging
Solution Approach 1:
The delivery path is segmented into a sequence of cross-sectional shapes at different heights. Each cross-section is represented as a separate geometric shape (polygon, circle, or ellipse) with defined vertices or control points. This segmentation transforms the complex 3D path extraction problem into a series of simpler 2D shape definitions at discrete height levels, making the encoding process more manageable and systematic
Solution Approach 2:
The patent transitions from representing delivery paths in traditional 2D top-down view to a 3D vertical cross-sectional representation. By defining paths as sequences of cross-sectional shapes with height information, the system adds the vertical dimension to path representation, enabling more accurate depiction of variable width paths and obstacle avoidance in three-dimensional space
2Reliability
If UAVs navigate using existing detection methods, then delivery can be attempted, but detecting delivery surfaces and obstacles remains challenging
Solution Approach 1:
The system performs preliminary detection and mapping of delivery surfaces and obstacles before UAV navigation. By pre-processing the environment data to create detailed 3D models of delivery locations and identifying all obstacles in advance, the system prepares comprehensive path information that guides the UAV, reducing the detection burden during actual flight and improving reliability
Solution Approach 2:
The patent introduces an intermediate data structure consisting of cross-sectional shapes that mediates between raw sensor data and UAV navigation commands. This intermediate representation serves as a bridge, translating complex LIDAR and aerial imagery data into simplified geometric forms that are easier for the UAV to process and follow, thereby improving detection reliability
3Adaptability or versatility
If variable width paths are used for UAV delivery, then obstacle avoidance is improved, but path representation and encoding becomes more complex
Solution Approach 1:
The patent implements dynamic path representation by allowing the cross-sectional shapes to vary in size, shape, and position along the vertical height of the delivery path. This dynamic approach enables the path to adapt to obstacles and terrain variations, with each cross-section independently defined to reflect the actual available space at that height, providing versatility for obstacle avoidance
Solution Approach 2:
The system changes parameters of the cross-sectional shapes (vertices, control points, radii) at different heights to represent variable width paths. By modifying these geometric parameters along the vertical axis, the system can accurately depict paths that narrow or widen to avoid obstacles, maintaining adaptability while using a consistent mathematical representation
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An approach is provided for constructing a delivery path that enables a UAV to safely access a delivery surface and avoids restricted access surfaces from the open sky. The approach involves detennining at least one delivery path to at least one delivery surface, wherein the delivery path represents at least one three-dimensional variable width path along which an aerial delivery vehicle can access the at least one delivery surface. The approach also involves transecting the delivery path with one or more planar surfaces. The approach further involves determining one or more shapes on the one or more planar surfaces, wherein the one or more shapes represent one or more intersections of the delivery path and the one or more planar surfaces. The approach also involves constructing at least one delivery path data object comprising at least one ordered list of the one or more shapes to represent the delivery path.