Performance and Biomechanical Analysis

horse and rider arena eventing jumping brush spread with machine learning and ai overlay
pony and rider dressage test at royal windsor horse show with ai and machine learning overlay

EPA completes Biomechanical analysis as part of the overall Performance Analysis service offered. This allows for greater scientific detail, therefore greater improvements in training from the comprehensive feedback.

Biomechanical analysis is completed by our AI programme, which calculates and compares rider's joint angles and symmetry during videos. The difference in biomechanics at key points in performance may indicate the possibility of an error occurring. Identifying areas of improvement can help increase training specificity.

The combination of the rider's biomechanical analysis and performance analysis can help distinguish potential cause and effect patterns in training and competition videos. Identification of rider's asymmetry and relation to decreased performance quality can also help improve specificity in training, both on and off horse.

For example, if the rider's symmetrical measurements and biomechanics are similar when landing in balance after a fence, then an anomalous measurement occurring when landing unbalanced, therefore unprepared for the next fence, could help identify an area of improvement for training. 

All joint angles and measurements are calculated during a test or jumping round in order to accurately quantify the analysis.

An example screenshot from the video

horse and rider dressage sparsholt college in trot, showing ai and machine learning overlay with numerical, chart and graphical data

Video with the skeleton overlay. The left side is blue and the right side is red - corresponding with the angles written underneath.

The left and right side shoulder, elbow, hip, knee and ankle angles. The angles change constantly throughout the video as the AI adjusts and remeasures the rider.

The graph measures the movement of the rider's centre of mass throughout the video. This indicates when there may be too much vertical movement, or alternatively too little - potentially showing the rider may have become more tense.

The bar chart shows the average angles of the elbow, hip, knee and ankle, on both right and left sides of the rider, throughout the video. These results indicate potential discrepancy of the rider's left and right-side angles which may cause an asymmetric position and result in performance or physiology issues.

trot pic for com graph.png
horse and rider dressage sparsholt college in walk, showing ai and machine learning overlay with numerical, chart and graphical data

This is an additional example screenshot where the skeleton overlay can be seen

over the horses ears. This is not where the rider is sitting,but it is where the AI draws the lines for the hip angles.

trot pic for com graph.png

In this example the red lines denoting the  right leg can be seen over the horses shoulder. The AI deduces where the right leg is and can therefore deduce the relevant angles. The red line is shown over the shoulder because this is where the leg is and would be seen by the camera on this angle if the horse were to be made transparent!

horse and rider dressage sparsholt college in trot, showing ai and machine learning overlay with numerical, chart and graphical data
trot pic for com graph.png