Proof of concept of a wearable IoT-based system for monitoring respiratory rate and surface temperature in horses.

Overview

• This research developed a wearable Internet of Things (IoT) system that non-invasively measures the respiratory rate and coat surface temperature of horses in real time and demonstrated that the device provides accurate and reliable data comparable to traditional manual methods.

Objective and Purpose

• The primary objective was to design and validate a wearable device capable of continuously monitoring two important physiological parameters in horses: respiratory rate (RR) and coat surface temperature (CST).
• This aims to improve real-time animal health monitoring to enhance welfare and detect early signs of illness or stress.

Device Design and Technology

• The device uses an embedded system with an ESP-32 microcontroller, a popular choice for IoT projects due to its wireless connectivity and low power consumption.
• Sensors include:
  • A strain gauge sensor to measure respiratory rate by detecting thoracic movements associated with breathing.
  • A DS18B20 temperature sensor to measure the surface temperature of the horse’s coat.
• Data are wirelessly transmitted in real time via Wi-Fi to allow immediate monitoring.
• Data are also stored locally on SD cards as a backup to ensure no data loss even in cases of connectivity issues.

Validation Methods

• The system’s measurements were validated by comparing data collected by the device to those obtained through conventional manual measurement methods under field conditions.
• Statistical methods used to assess agreement and reproducibility:
  • Intraclass Correlation Coefficient (ICC) evaluated how well the device’s measurements agreed with manual measurements, indicating reliability and reproducibility.
  • Bland-Altman plots analyzed the bias (systematic difference) and limits of agreement between the two measurement methods.
  • Paired t-tests checked if there were statistically significant differences in measurements from the two methods.

Results

• The device showed excellent reproducibility for respiratory rate with an ICC of 0.85, indicating strong agreement with manual measurements.
• The coat surface temperature showed very good reproducibility with an ICC of 0.67, reflecting a strong but slightly lower agreement than respiratory rate.
• No significant differences were found between the device and manual methods (p > 0.05 in paired t-tests), supporting the device’s accuracy.
• Bland-Altman analysis revealed low bias and narrow limits of agreement, meaning the device’s readings closely matched conventional methods without systematic errors or large discrepancies.

Implications and Conclusions

• This proof-of-concept confirms that the wearable IoT system can reliably and accurately measure respiratory rate and coat surface temperature in horses.
• The non-invasive nature and real-time wireless data transmission make this technology highly suitable for continuous health monitoring in animals.
• Such a device could have practical applications in veterinary care, performance monitoring, and welfare assessment by enabling early detection of respiratory or thermal stress conditions.
• Future work could focus on expanding parameters measured, improving sensor integration, scaling to more animals, and integrating with broader animal health management systems.