Abstract:Drones may be more advantageous than fixed cameras for quality control applications in industrial facilities, since they can be redeployed dynamically and adjusted to production planning. The practical scenario that has motivated this paper, image acquisition with drones in a car manufacturing plant, requires drone positioning accuracy in the order of 5 cm. During repetitive manufacturing processes, it is assumed that quality control imaging drones will follow highly deterministic periodic paths, stop at predefined points to take images and send them to image recognition servers. Therefore, by relying on prior knowledge about production chain schedules, it is possible to optimize the positioning technologies for the drones to stay at all times within the boundaries of their flight plans, which will be composed of stopping points and the paths in between. This involves mitigating issues such as temporary blocking of line-of-sight between the drone and any existing radio beacons; sensor data noise; and the loss of visual references. We present a self-corrective solution for this purpose. It corrects visual odometer readings based on filtered and clustered Ultra-Wide Band (UWB) data, as an alternative to direct Kalman fusion. The approach combines the advantages of these technologies when at least one of them works properly at any measurement spot. It has three method components: independent Kalman filtering, data association by means of stream clustering and mutual correction of sensor readings based on the generation of cumulative correction vectors. The approach is inspired by the observation that UWB positioning works reasonably well at static spots whereas visual odometer measurements reflect straight displacements correctly but can underestimate their length. Our experimental results demonstrate the advantages of the approach in the application scenario over Kalman fusion.

What problem does this paper attempt to address?

This paper mainly explores the issue of precise positioning when using drones for quality control in indoor facilities. In current industrial applications, drones have higher flexibility compared to fixed cameras and can be dynamically redeployed according to production plans. In environments such as car manufacturing plants, the positioning accuracy of drones needs to reach 5 centimeters. Drones move along predefined paths during repetitive production processes, stop at specific points to capture images, and send them to image recognition servers. A self-calibrating sensor fusion solution is proposed in the study to address issues such as line of sight occlusion, sensor data noise, and visual reference loss. This method is based on filtering and clustering of Ultra-Wideband (UWB) data to calibrate visual odometry readings as an alternative to direct Kalman fusion. This approach combines the advantages of both techniques when at least one of them is functioning properly. Experimental results show that this method outperforms Kalman fusion in stop point detection and trajectory estimation errors. The background section of the paper introduces the applications of drones in industrial scenarios, such as monitoring, just-in-time part delivery, and inventory control, and discusses the positioning challenges of drones in indoor environments. Visual odometry and UWB positioning were chosen as the main positioning technologies in the research because they require fewer reference points, are lightweight, and can be easily deployed in technical environments. The hardware design section describes the experimental setup using the Pozyx positioning system and Intel RealSense T265 tracking camera, as well as how data is extracted from these two sensors. The proposed self-calibration method includes independent Kalman filtering, data association through flow clustering, and mutual calibration of sensor readings based on accumulated correction values. Experimental results demonstrate that this approach has advantages over Kalman fusion in specific application scenarios, especially in stop point detection and trajectory estimation accuracy. In conclusion, this paper aims to address the issue of precise positioning of drones in indoor environments and proposes a self-calibrating sensor fusion strategy using UWB and visual odometry to improve positioning performance.

Self-Corrective Sensor Fusion for Drone Positioning in Indoor Facilities

Indoor Wireless Positioning Utilizing WSN and Machine Vision

A Drone Localization System Based On Multiple Time Delays Fusion

Indoor Drone Positioning: Accuracy and Cost Trade-Off for Sensor Fusion

Position Estimation Method for Small Drones Based on the Fusion of Multisource, Multimodal Data and Digital Twins

UVIO: Adaptive Kalman Filtering UWB-Aided Visual-Inertial SLAM System for Complex Indoor Environments

Data-Driven Indoor Positioning Correction for Infrastructure-Enabled Autonomous Driving Systems: A Lifelong Framework

A Novel Loosely Coupling Fusion Approach of Ultra-Wideband and Wheel Odometry for Indoor Localisation

A Novel Positioning Module and Fusion Algorithm for Unmanned Aerial Vehicle Monitoring

Onboard Real-Time Multi-Sensor Pose Estimation for Indoor Quadrotor Navigation with Intermittent Communication

Adaptive Extended Kalman Filter-Based Fusion Approach for High-Precision UAV Positioning in Extremely Confined Environments

Airborne sensor fusion: Expected accuracy and behavior of a concurrent adjustment

An Indoor Positioning Method Based on UWB and Visual Fusion

Research on a Visual/Ultra-Wideband Tightly Coupled Fusion Localization Algorithm

Delayed Monocular SLAM Approach Applied to Unmanned Aerial Vehicles

Constrained ESKF for UAV Positioning in Indoor Corridor Environment Based on IMU and WiFi

Real-Time Onboard 3D State Estimation of an Unmanned Aerial Vehicle in Multi-Environments Using Multi-Sensor Data Fusion

Vision-Based UAV Detection and Localization to Indoor Positioning System

Kalman-Filter-Based Integration of IMU and UWB for High-Accuracy Indoor Positioning and Navigation