Multi-Sensor Fusion for Robust MAV Navigation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing autonomous flight technologies for rotorcraft micro-aerial vehicles (MAVs) face challenges in indoor and outdoor environments due to reliance on single sensor modalities, which fail in environments with magnetic interference or lack of GPS, leading to unreliable navigation and abrupt changes in velocity and trajectory during sensor transitions.
Innovation Solution
A modular and extensible approach integrating noisy measurements from multiple heterogeneous sensors, using an Unscented Kalman Filter (UKF) for state estimation and control, allowing for smooth and globally consistent position estimates in real-time, even in environments with varying sensor availability and quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If GPS-based autonomous flight is used in outdoor environments, then navigation accuracy is improved, but the system fails in indoor environments or urban canyons where GPS signals are unavailable
Solution Approach 1:
The patent combines multiple sensor modalities (GPS, magnetometer, barometer, inertial sensors, and visual odometry) into a unified navigation system using the Factor Graph framework. This merging allows the system to leverage the strengths of each sensor while compensating for their individual weaknesses, enabling operation across both outdoor GPS-available and indoor GPS-denied environments.
Solution Approach 2:
The navigation system is designed to be universally applicable across different environments by implementing multiple sensor modalities that can function independently or in combination. The system can operate using GPS outdoors, switch to magnetometer and barometer in transitional zones, and rely on visual odometry and inertial sensors indoors, making it multi-functional across diverse operational contexts.
2Measurement precision
If magnetometer is used for navigation in indoor environments, then position estimation is improved, but the sensor becomes unreliable due to magnetic interference from building structures
Solution Approach 1:
The patent introduces visual odometry as an intermediary measurement modality that mediates between the magnetometer and the overall navigation system. When magnetic interference degrades magnetometer reliability, the visual odometry system provides alternative position estimates that are fused with magnetometer data through the Factor Graph framework, maintaining navigation accuracy without being affected by magnetic interference.
Solution Approach 2:
The system dynamically changes the weighting and reliability parameters of different sensors based on environmental conditions. When magnetic interference is detected or suspected in indoor environments, the Factor Graph framework automatically adjusts the confidence weights assigned to magnetometer measurements versus visual odometry measurements, effectively adapting to changing sensor reliability without manual intervention.
3Device complexity
If single sensor modality is used for autonomous flight, then system complexity is reduced, but the system becomes unreliable when sensor assumptions fail in certain environments
Solution Approach 1:
The patent segments the navigation system into independent modular components (GPS module, magnetometer module, barometer module, visual odometry module, inertial navigation module) that can be individually developed, tested, and maintained. Each module processes its own sensor data independently before fusion, which manages complexity through modular architecture while achieving high reliability through diversity of sensor modalities.
4Measurement precision
If abrupt correction is applied when GPS becomes available after indoor flight, then position error is corrected quickly, but the transition causes large changes in velocity and trajectory that disrupt smooth flight
Solution Approach 1:
The Factor Graph framework performs beforehand cushioning by continuously maintaining a consistent probabilistic state estimate that incorporates all available sensor measurements. When GPS becomes available after indoor flight, the framework has already been accumulating visual odometry and inertial navigation data, so the transition to GPS-aided navigation is smooth and continuous, avoiding abrupt corrections that would disrupt flight stability.
Data Source
AI summary
The subject matter described herein includes a modular and extensible approach to integrate noisy measurements from multiple heterogeneous sensors that yield either absolute or relative observations at different and varying time intervals, and to provide smooth and globally consistent estimates of position in real time for autonomous flight. We describe the development of the algorithms and software architecture for a new 1.9 kg MAV platform equipped with an IMU, laser scanner, stereo cameras, pressure altimeter, magnetometer, and a GPS receiver, in which the state estimation and control are performed onboard on an Intel NUC 3rd generation i3 processor. We illustrate the robustness of our framework in large-scale, indoor-outdoor autonomous aerial navigation experiments involving traversals of over 440 meters at average speeds of 1.5 m/s with winds around 10 mph while entering and exiting buildings.