IoT Smart Wearable for Horses: WHOOP-Style Equine Health System

Developing a 24/7 wearable health monitoring system for horses that will be used regularly under a saddle brings with it numerous biomechanical, electrical, and environmental challenges that need to be accounted for.

Unlike devices designed for use by humans, equine wearables need to accommodate high impact forces from movement, constant friction against a surface, continual disconnection of signals due to sweat, and variable exposure to the elements.

This project will take a 10cm working prototype at an early stage of advancement and turn it into an equine wearable system that is clinically reliable as well as ready for mass production.

The goal of this project is to enable real-time health and physiological data, early disease detection, and performance insight through the use of medical grade sensors, smart textiles, and edge AI.

To accomplish a more streamlined and ergonomic shape (5-8 mm), a flex-rigid PCB architecture was residential, in order to replace rigid boards.

This PCB design method is intended to improve durability of hardware, reduce mechanical failure points, and improve ride's comfort under saddle pressure.

The main processing unit of the wearables uses the Nordic nRF52840 / nRF5340 microcontroller (>2). These processors provide ultra-low power processing, as well as Bluetooth low energy (BLE v5.3) functionality.

The next generation processors provide additional functionality such as directed advertising, and long range communication, creating seamless communication between the wearable, the rider's mobile phone, and the stable gateway.

The power management of the system maximises the useable capacity of the rechargeable battery (Li-Po pouch), and uses Qi wireless charging (eliminating the need for an external charging port) resulting in improved waterproofing (with an intention to achieve an IP68 waterproof rating), and resilience to environmental condition such as sweat, mud and rain.

Thermal management, and EMI shielding, were both designed into the wearable from the outset such that signals remain stable and long-lasting during an extended operation period.

A critical factor distinguishing this system from other systems is its ability to combine multiple sensors to create relevant health data (metrics).

Cardiovascular Monitoring (ECG with PPG)

This system is designed to produce ECG data by using a conductive textile (cloth) electrode to record a single lead ECG signal.

Therefore, through the use of an AFE (Analog Front End), namely MAX30001, this system is able to accurately acquire ECG and bio-impedance information. From that data, the system can derive the following information:

• Heart Rate
• Heart Rate Variability
• Respiratory Rate (via impedance)

Monitoring Motion and Behavior

All motion patterns for the horse are continuously monitored via a 6-axis IMU (i.e., Bosch BMI270) which provides the ability to identify:

• Gait abnormalities
• Restlessness or pawing
• Early signs of conditions such as colic

Thermal Monitoring

To obtain an estimate of a horse's core body temperature, this system utilizes a dual-sensor thermal system consisting of:

• Skin-contact sensor
• Ambient-facing sensor

The temperature estimation is made more accurate beyond that of skin surface readings using heat flux modeling.

The chest band is not merely a wearable accessory , it is an active sensing platform.

Conductive Materials

Silver-coated and carbon-infused yarns are woven directly into the textile to function as ECG electrodes. These materials are selected for:

• High conductivity
• Durability under repeated stress
• Resistance to sweat corrosion

Compression Optimization

Maintaining consistent skin contact is critical for signal accuracy. The band is engineered with a defined compression modulus, ensuring:

• Stable electrode contact during movement
• No restriction of breathing or lung expansion

Encapsulation & Comfort

The electronics module is encapsulated using medical-grade silicone or TPU, minimizing pressure points and preventing skin irritation or "girth itch". Long-term wearability and animal comfort are prioritized to meet veterinary standards.

The final result of the system is to convert the collected additional sensor data to meet immediate results.

Recovery Analysis Using HRV

The recovery assessment (using values like RMSSD) is based on time-domain metrics to evaluate activity level regarding the parasympathetic nervous system (PSNS); the following provides an example of data collected for each user:

• The user's current recovery level is on a scale from one to ten
• The user will also receive a rating for their stress level (based on RMSSD)
• The user will receive an assessment of their risk of overtraining

Early Detection System for Colic

Colic continues to be the most common Causes of death in the horse. By utilizing data collected from heart rate and patterns of IMU, machine learning models (both Random Forest and CNN) can be developed to identify early warning signs of colic as follows:

• Elevated Heart rate
• Unusual rolling behavior
• Repeated flank observation

Edge Computing Plan

To conserve battery life and reduce bandwidth consumption, edge computing is utilized on the MCU prior to the transmission of any data to the cloud. Therefore, rather than sending raw data continuously, only the following is sent to the cloud:

• Processed data summary (i.e. heart rate, respiration)
• Notifications to the user based on any of the above events

This allows for effective real time monitoring while consuming minimal power.

A structured development roadmap ensures technical validation and commercial readiness:

Phase 1: Proof of Concept (Weeks 1-4)

• Validate ECG signal quality using textile electrodes
• Test AFE performance in real-world conditions

Phase 2: Alpha Prototype (Weeks 5-12)

• Develop flex-PCB and compact housing
• Implement baseline algorithms for activity detection

Phase 3: Pilot Deployment (Months 3-6)

• Deploy ~50 units in professional stables
• Collect labeled datasets for model training (colic, stress, recovery)

Phase 4: Certification & Manufacturing (Months 6-12)

• Finalize Design for Manufacturing (DFM)
• Conduct RF, safety, and durability testing
• Initiate scalable production

Creating a horse equivalent of the "WHOOP" is not simply a matter of applying human wearables, an entire process must be created from scratch using electronics, fabric science, and input from veterinarians.

This system is a combination of strong hardware engineering, intelligent fabric integration, and AI-supported data analysis, resulting in a system that can ultimately provide equine health tracking with a viable product that has the ability to scale and provide clinically useful data.

As this project moves from prototype to mass production, there will be a greater emphasis on reliability, compliance with necessary regulations, and validation in the real world to ensure that it functions in both controlled environments and through the rigors of everyday equine usage.