Perceptive AV Navigation Using Asset Feature Coordinate Frames
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles (AVs) face challenges in navigating accurately without GPS, particularly in GPS-denied environments, such as underground, indoors, or under bridges, where traditional GPS-based location methods are insufficient for civilian inspection applications requiring high precision.
Innovation Solution
The system employs machine perception using a combination of sensors like cameras, LIDAR, and computer vision to create a perceptive navigation system that identifies asset features, allowing AVs to determine their position and orientation relative to these features, enabling autonomous navigation without relying on GPS. This involves a system controller, localization module, and environmental sensors that detect and classify asset features, transforming the AV's pose into a coordinate system associated with these features.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If GPS-based location determination is used, then navigation simplicity is maintained, but location accuracy deteriorates in GPS-denied environments
Solution Approach 1:
The system segments the navigation problem into multiple components: GPS-based navigation for open environments and vision-based navigation for GPS-denied environments. The GPS receiver provides coarse location information when available, while the vision system with feature detection and SLAM algorithms provides fine-grained localization when GPS is unavailable, allowing the system to adapt to different environmental conditions
Solution Approach 2:
The patent introduces visual features and artificial markers as intermediary objects that bridge the gap between the vehicle and the environment for localization. These features serve as reference points that the vision system can detect and use to calculate the vehicle's position and orientation, enabling accurate navigation without direct GPS signals
2Measurement precision
If high-resolution imaging is performed, then inspection quality improves, but proximity to targets is required increasing navigation complexity
Solution Approach 1:
The system replaces complex mechanical positioning systems with a vision-based navigation approach. Instead of using sophisticated mechanical actuators and sensors to achieve precise positioning, the system uses computer vision algorithms to detect visual features and calculate the vehicle's position and orientation, simplifying the mechanical requirements while maintaining high inspection quality
Solution Approach 2:
The patent transitions from two-dimensional image capture to three-dimensional spatial understanding by incorporating depth information through stereo vision or time-of-flight sensors. This dimensional enhancement allows the system to not only capture high-resolution images but also determine the vehicle's precise position and orientation in 3D space, enabling accurate navigation and inspection
3Loss of information
If wireless data streaming is implemented, then real-time monitoring is achieved, but data transmission dependency increases
Solution Approach 1:
The system implements self-service by processing and analyzing data locally on the autonomous vehicle using onboard computers and processors. The vehicle can independently perform navigation decisions, obstacle detection, and inspection data analysis without requiring continuous wireless communication, reducing dependency on external data transmission while maintaining operational reliability
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
The system achieves accurate navigation within 2 meters, comparable to human inspectors, by using onboard-generated image data and sensor information, allowing AVs to operate autonomously and adapt mission plans in real-time, even in environments where GPS is unreliable.
Implementation Method 1
The system employs machine perception using a combination of sensors like cameras, LIDAR, and computer vision
Data Source
AI summary
An agricultural navigation system and method for an autonomous vehicle (AV) is described. The agricultural navigation system includes a system controller, a localization module associated with the system controller, and an environmental sensor. The system controller determines an AV positional pose that identifies the location of the AV. The system controller determines a relative body frame of reference (RBF) that is associated with the AV positional pose. The environmental sensor detects an asset feature in the AV environment. The asset feature includes an agricultural asset feature having a crop row. The system controller identifies at least one asset feature frame (AFF) that includes a coordinate system originating at the asset feature. The localization module determines the AV positional pose in the coordinate system of the AFF. The system controller transforms the AV positional pose from the RBF coordinate system to the coordinate system of the AFF.