Thumbnail Image
Publication

Advancing livestock welfare sensors: an optimization-based BLE & IMU localization approach

Sobczak, Eric J.
Citations
Altmetric:
Advisor
Vincent, Tyrone
Editor
Date
Date Issued
2025
Date Submitted
Keywords
animal welfare
 Bluetooth
 IMU
 localization
Collections
2025 - Mines Theses & Dissertations
Show all metadata
Files
Thumbnail Image
Sobczak_mines_0052N_13050.pdf
Adobe PDF, 33.95 MB
Research Projects
Organizational Units
Journal Issue
URI
https://hdl.handle.net/11124/181130
Embargo Expires
Abstract
With the expanding concept of precision livestock farming (PLF), an increasing number of animals wear welfare sensors equipped with Bluetooth Low Energy (BLE) and Inertial Measurement Unit (IMU) capabilities. Accurate positioning information from these sensors provides valuable insights into livestock welfare and behavior. Traditional Global Navigation Satellite System (GNSS) sensors are bulky, costly, and perform poorly indoors, whereas existing BLE solutions typically require complex calibration procedures and rely on restrictive environmental assumptions. This study proposes an algorithm utilizing optimization techniques and advancements in the Bluetooth communication standard to achieve RSSI-based localization. A purpose-built welfare sensor was developed, functioning as both reader and beacon, enabling inter-device communication and data collection. By combining BLE RSSI measurements with IMU data and applying a Kalman filter for sample noise reduction, the algorithm successfully achieved sub 1 meter accuracy without requiring gate analysis or fingerprinting techniques. An ArUco-based camera ground truth system was established to evaluate the system performance. While the integrated IMU sensor did not directly enhance positional estimation accuracy, the processed velocity norm effectively indicated device movement and was used for dynamically tuning Kalman filter gains. Additionally, employing optimization during the final measurement-solving step allowed for weighted prioritization of accurate 1D distance estimates, further enhancing localization accuracy and consistency. These combined improvements demonstrate the viability of this optimized localization technique for practical livestock monitoring applications.
Associated Publications
Rights
Copyright of the original work is retained by the author.
Embedded videos