3D Collision Probability Mapping for Dynamic Aerial Obstacles
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Aerial vehicles face challenges in navigating complex 3D environments due to static and dynamic obstacles, leading to safety concerns and limitations in autonomous operation, as existing routing methods rely on onboard sensors or human intervention and lack accurate collision probability assessments.
Innovation Solution
A system that partitions 3D space into varying resolution shapes, monitors aerial vehicle flights, records dynamic obstacle observations, and transmits data to calculate collision probability maps, enabling safer route planning by predicting dynamic obstacle density and integrating this data into a collision-probability 3D map for route computation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If aerial vehicles use onboard sensors or human intervention for routing, then navigation can be performed, but safety and collision probability assessment are insufficient
Solution Approach 1:
The system pre-calculates collision probability maps for 3D space partitions before aerial vehicles operate. By computing collision probabilities in advance based on historical obstacle data and environmental factors, the system provides reliable safety assessments that guide route planning, eliminating the need for reactive onboard sensor detection during critical moments.
Solution Approach 2:
The patent introduces a server-based collision probability calculation system as an intermediary between aerial vehicles and the 3D environment. This intermediary processes environmental data, computes collision probabilities for space partitions, and returns assessment results to vehicles, providing sophisticated safety evaluation that neither simple onboard sensors nor human operators can achieve alone.
2Measurement precision
If 3D space is partitioned into fine-resolution shapes for accurate collision assessment, then measurement precision improves, but device complexity and computational load increase
Solution Approach 1:
The 3D operating space is divided into discrete partition units with associated collision probability values. By segmenting the continuous space into manageable cells, the system achieves precise collision assessment without requiring complex continuous mathematical models. Each partition can be independently calculated and stored, simplifying the overall system architecture.
Solution Approach 2:
Instead of performing complex real-time calculations onboard each aerial vehicle, the system pre-computes collision probability maps and stores them as data structures. Vehicles copy these pre-calculated probability values for their respective space partitions, avoiding redundant computations and reducing onboard processing complexity while maintaining high measurement precision.
3Measurement precision
If real-time monitoring and recording of dynamic obstacles is implemented, then collision probability data accuracy improves, but loss of time and processing overhead increase
Solution Approach 1:
The system implements periodic sampling of aerial vehicle positions and obstacle data rather than continuous monitoring. By recording dynamic obstacle information at regular intervals and updating collision probability maps periodically, the system maintains accurate real-time awareness while reducing processing overhead and time losses associated with constant data collection and computation.
Data Source
AI summary
An approach is provided for dynamic obstacle data in a collision probability map. The approach, for example, involves monitoring a flight of an aerial vehicle through a three-dimensional (3D) space that is partitioned into 3D shapes of varying resolutions. The approach also involves detecting an entry of the aerial vehicle into one 3D shape of the plurality of 3D shapes. The approach further involves, on detecting an exit of the aerial vehicle form the one 3D shape, recording a 3D shape identifier (ID) of the one 3D shape and at least one of a first timestamp indicating the entry, a second timestamp indicating the exit, a duration of stay in the one 3D shape, dimensions of the aerial vehicle, or a combination thereof as a dynamic obstacle observation record. The approach further involves transmitting the dynamic obstacle observation record to another device (e.g., a server for creating the collision probability map).