Automatic methods of hoof-on and -off detection in horses using wearable inertial sensors during walk and trot on asphalt, sand and grass.
Authors: Briggs Eloise V, Mazzà Claudia
Journal: PloS one
Summary
# Editorial Summary: Hoof-on and -off Detection Using Wearable Inertial Sensors Accurate identification of hoof contact and lift-off events forms the foundation of equine gait analysis, yet traditional force plate systems remain impractical for field-based assessment. Briggs and Mazza (2021) tested wearable inertial measurement units mounted at three locations—hoof, pastern and cannon—across eleven horses of mixed breeds and ages, recording walk and trot on asphalt, grass and sand surfaces to determine which sensor placement could most reliably detect gait events. Pastern-mounted sensors proved most accurate, achieving mean detection errors of just 1–6 milliseconds on asphalt depending on limb and gait, whilst maintaining consistent accuracy across all surfaces with stance duration errors of only 0.1–1.16% of stride cycle. These findings are significant for practitioners because they validate pastern-mounted inertial sensors as a practical alternative to force plates, enabling longitudinal lameness monitoring, gait classification and rehabilitation tracking in clinical and field settings without the constraints of laboratory equipment. For farriers, veterinarians and rehabilitation specialists, this methodology opens possibilities for objective, real-time gait assessment during routine work and training.
Read the full abstract on PubMed
Practical Takeaways
- •Pastern-mounted inertial sensors offer a portable, field-friendly alternative to force plates for monitoring hoof timing and stance duration during training and rehabilitation
- •These sensors maintain accuracy across different surfaces (asphalt, grass, sand), making them practical for real-world farm and exercise situations
- •The sub-1% error in stance phase measurement means pastern sensors are reliable enough for detecting lameness-related gait asymmetries in clinical and performance contexts
Key Findings
- •Pastern-mounted inertial sensors achieved mean errors of 1-6ms for gait event detection on asphalt across different limbs and gaits
- •Pastern-based methods showed consistent errors of 0.1-1.16% of stride cycle when measuring stance duration across all surface types (asphalt, grass, sand)
- •Sensor placement at pastern level outperformed cannon-mounted sensors for gait event detection accuracy and precision
- •Validated methods enable reliable equine gait event detection under field conditions on multiple exercise surfaces