3D Aerial Route Encoding With Hierarchical Collision Maps
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Aerial vehicles face challenges in navigating complex 3D environments due to the absence of proper 3D modeling, relying on onboard sensors or human operators for obstacle detection, which limits their safe operation and autonomous adoption.
Innovation Solution
A system that calculates aerial routes using a pre-computed collision-probability 3D map, partitioning 3D space into a hierarchy of shapes and generating a traversal command list to encode and decode routes for safe navigation, incorporating collision probability data to preempt potential risks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If aerial vehicles rely on onboard sensors or human operators for obstacle detection, then the system complexity is reduced, but the safety and reliability of navigation in complex 3D environments deteriorates
Solution Approach 1:
The patent applies preliminary action by pre-computing collision probability data and creating 3D models of the environment before aerial vehicle operation. The system prepares route information, obstacle data, and collision probabilities in advance, allowing the vehicle to navigate safely without complex onboard detection systems. This shifts the computational burden from the vehicle to the ground system.
Solution Approach 2:
The patent introduces an intermediary 3D collision probability map that mediates between the aerial vehicle and the complex environment. This digital model serves as a simplified representation that the vehicle can use for safe navigation without directly interacting with the full complexity of the physical environment. The intermediary model translates complex environmental data into usable route guidance.
2Extent of automation
If detailed 3D modeling and collision probability data are provided for route planning, then the reliability of autonomous navigation is improved, but the device complexity and data processing requirements worsen
Solution Approach 1:
The patent applies segmentation by dividing the complex 3D environment into discrete volumetric elements or grid cells. Each element can be independently analyzed for collision probability, allowing the system to manage complexity through modular processing. This segmentation enables efficient computation and storage of collision data without requiring complete environmental modeling.
Solution Approach 2:
The patent creates a simplified digital copy or representation of the physical environment in the form of a 3D collision probability map. This virtual model captures essential navigation information without replicating the full complexity of the physical world. The copy allows autonomous navigation decisions to be made based on processed data rather than raw environmental complexity.
3Loss of information
If aerial vehicles operate in complex 3D spaces without proper 3D modeling, then the ease of operation is maintained, but the loss of information regarding collision risks increases
Solution Approach 1:
The patent transitions from traditional 2D map representations to 3D volumetric modeling to capture collision probability information. By adding the vertical dimension and volumetric depth, the system can represent obstacles and safe zones in three-dimensional space, preserving critical navigation information that would be lost in flat 2D representations. This dimensional enhancement enables accurate modeling of aerial environments.
Data Source
AI summary
An approach is provided for representing an aerial route in a three-dimensional (3D) space. The approach, for example, involves determining a root 3D shape encompassing an aerial route through a 3D space. The 3D space is partitioned into a hierarchy of 3D shapes, and the aerial route comprises a list of 3D shape identifiers (IDs) corresponding to the 3D shapes that form the aerial route. The approach also involves recursively comparing a respective 3D shape ID of the root 3D shape or child shapes of the root 3D shape to the list of 3D shape IDs of the aerial route. The approach further involves generating a traversal command list by selectively adding the respective 3D shape ID that matches a 3D shape ID of the aerial route or includes the 3D shape ID as a child 3D shape ID. The approach further involves providing the traversal command list as an output.