UAV 360-Degree Obstacle Avoidance Using Triangulated Laser Lines
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current sensors for autonomous mobility and mapping on small aerial platforms, such as quadrotors, face limitations due to size, weight, and power constraints, with existing LADAR options being scarce and stereo vision or structure from motion methods being ineffective in poor lighting conditions.
Innovation Solution
The use of multiple fisheye cameras and laser line scanners with a near-infrared laser projection system, creating vertical lines that are triangulated in 3D space to provide full 360-degree range sensing capabilities, allowing for accurate obstacle avoidance and mapping even in featureless environments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional LADAR sensors are used for autonomous mobility and mapping, then measurement precision is improved, but device complexity and weight increase beyond SWAP constraints
Solution Approach 1:
The patent segments the LADAR sensing function into multiple simpler line sensors arranged in a specific geometric configuration. Instead of using a single complex LADAR sensor, multiple line sensors are distributed across the UAV platform, with each sensor providing a portion of the overall 3D mapping capability. This segmentation reduces individual sensor complexity while maintaining collective measurement precision.
Solution Approach 2:
The patent combines multiple line sensor measurements into a unified 3D environmental model. By merging data from multiple simplified sensors positioned at different locations and orientations, the system achieves comprehensive spatial awareness equivalent to or exceeding traditional LADAR, while keeping individual sensor units simple and lightweight.
2Device complexity
If a sensor with small vertical field of view is installed horizontally on the quadrotor, then device complexity is reduced, but the vehicle becomes blind in the direction of travel at high speeds due to pitch
Solution Approach 1:
The patent employs an asymmetric arrangement of line sensors relative to the quadrotor's center of gravity and flight path. Sensors are positioned and oriented at specific angles that compensate for the vehicle's pitch motion during high-speed travel. This asymmetric configuration ensures that the combined field of view of all sensors maintains continuous coverage of the direction of travel despite pitch variations up to 45 degrees.
Solution Approach 2:
The patent adds the temporal dimension to the sensor configuration, where sensors capture data at multiple time points during pitch motion. By processing sequential measurements from the same horizontal sensors taken during pitch cycles, the system reconstructs three-dimensional spatial information that compensates for the limited instantaneous vertical field of view, effectively adding a time-based dimension to overcome the geometric limitation.
3Device complexity
If stereo vision or structure from motion is used, then device complexity is reduced, but measurement precision deteriorates in poor lighting conditions
Solution Approach 1:
The patent replaces passive optical systems (stereo vision, structure from motion) with an active illumination system. Instead of relying on ambient light to capture images for stereo matching or motion parallax, the system uses active LADAR sensors that emit their own light source. This substitution of passive optical mechanics with active electromagnetic illumination ensures measurement precision is maintained regardless of ambient lighting conditions, while keeping the overall system relatively simple.
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
This configuration enables high-density point cloud generation, accurate range measurement, and robust operation in various lighting conditions, meeting the unique needs of autonomous multicopters for both indoor and outdoor environments with minimal weight and moving parts.
Implementation Method 1
A near-infrared laser projection unit sends light out into the environment.
Implementation Method 2
If the light hits objects in the environment it is reflected and viewed by the cameras.
Data Source
AI summary
Structured light approaches utilize a laser to project features, which are then captured with a camera. By knowing the disparity between the laser emitter and the camera, the system can triangulate to find the range. Four, 185 degree field-of-view cameras provide overlapping views over nearly the whole unit sphere. The cameras are separated from each other to provide parallax. A near-infrared laser projection unit sends light out into the environment, which is reflected and viewed by the cameras. The laser projection system will create vertical lines, while the cameras will be displaced from each other horizontally. This relative shift of the lines, as viewed by different cameras, enables the lines to be triangulated in 3D space. At each point in time, a vertical stripe of the world will be triangulated. Over time, the laser line will be rotated over all yaw angles to provide full a 360 degree range.