Voxel-Based Path Collision Avoidance in Confined 3D Spaces
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for collision detection between vehicles and their environments in confined spaces are inefficient and lack accuracy, particularly when using real-time sensor data, as they require costly redesigns and are not feasible for design purposes.
Innovation Solution
A GPU-accelerated collision detection framework using voxelized representations of vehicles and point cloud data, employing a voxel-based Minkowski sum algorithm to compute collisions and clearance analysis, allowing for flexible design changes and accurate navigation path planning.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If real-time sensor data is used for collision detection, then navigation accuracy is improved, but redesign and remodeling costs increase
Solution Approach 1:
The patent creates a voxelized copy of the vehicle model and a point cloud representation of the environment. These digital copies enable collision detection without requiring physical sensors or repeated physical prototyping, thereby improving measurement precision while avoiding redesign costs.
Solution Approach 2:
The patent replaces physical sensor-based collision detection with a computational approach using voxelized models and point cloud data. This substitution eliminates the need for costly physical redesigns and remodeling while maintaining high collision detection accuracy through GPU-accelerated processing.
2Measurement precision
If voxel resolution is increased, then collision detection precision is improved, but computational complexity increases
Solution Approach 1:
The patent performs voxelization of the vehicle model and environment as a preliminary step before collision detection. This pre-processing creates ready-to-use voxel grids that can be reused across multiple collision checks, reducing computational complexity during actual navigation while maintaining high precision.
Solution Approach 2:
The patent transforms the continuous 3D space into a discrete voxel grid representation. This dimensional transformation enables efficient computational processing through GPU parallelization while maintaining high collision detection precision, effectively managing computational complexity through discretization.
3Speed
If GPU acceleration is used, then processing speed is improved, but hardware requirements increase
Solution Approach 1:
The patent introduces a voxelized representation as an intermediary between the vehicle model and collision detection process. This intermediary structure is optimized for GPU processing, enabling fast parallel computation while managing hardware requirements through efficient memory usage and data representation.
Data Source
AI summary
A path collision avoidance method and system may include obtaining a three-dimensional point cloud of an at least partially enclosed space, obtaining a voxelized model of a vehicle/robot, and outputting a visual representation of navigation of the vehicle/robot within the at least partially enclosed space based on the three-dimensional point cloud of the at least partially enclosed space and the voxelized model of the vehicle/robot.