Horse gait analysis using wearable inertial sensors and machine learning
Authors: Manju Rana, Vikas Mittal
Journal: Proceedings of the Institution of Mechanical Engineers, Part P: Journal of Sports Engineering and Technology
Summary
# Editorial Summary Rana and Mittal (2023) developed a wearable inertial measurement unit (IMU) system to address significant limitations in equine gait assessment, particularly the subjective nature of visual evaluation and the expense of optical motion capture systems. Their lightweight accelerometer-based sensors, mounted on each leg, captured ground reaction force (GRF) data and limb orientation during field work—allowing practitioners to move beyond the laboratory and assess horses during actual training and competition. Machine-learning classification models trained on sensor-derived features successfully distinguished between discrete movement patterns (walk, trot, canter, gallop, and jumping), with the implication that systematic lameness detection and performance monitoring become feasible outside traditional clinical settings. For farriers, veterinarians, and equine physiotherapists, this technology offers a practical, repeatable alternative to subjective gait assessment, potentially enabling earlier detection of subtle asymmetries and ground reaction force abnormalities before they manifest as overt lameness. The ability to trend kinetic and kinematic data over time during training cycles also supports evidence-based decisions about workload management, recovery protocols, and return-to-work timelines—making this tool particularly valuable for competition yards and rehabilitation programmes.
Read the full abstract on the publisher's site
Practical Takeaways
- •Wearable sensor technology offers an affordable, field-deployable option for monitoring horse gait and movement patterns without the expense and facility constraints of motion capture systems
- •Automated gait classification using machine learning could help identify lameness or movement abnormalities earlier than traditional visual assessment, supporting training and management decisions
- •This technology enables owners and riders to track quantitative gait data over time for evidence-based training adjustments and health monitoring
Key Findings
- •Wearable inertial measurement units (IMUs) provide a cost-effective alternative to optical motion capture for equine gait analysis
- •Machine learning models successfully classified horse movements including jumps, stands, gallops, and trots using accelerometer data
- •Ground Reaction Forces can be measured using 100g accelerometer data from each leg to detect movement anomalies