SLEIP AI -Computer Vision assessment of lameness in horses

Purpose and goal:

The project goal was to expand the technical capacity of our veterinary SaaS product by developing movement analysis of horses during lungeing. Our previous product was able to analyze horses in straight line trot. To provide a more complete diagnostic tool to our customers, we wanted to explore the possibility of performing this analysis also on the circle, a complex challenge for Computer Vision technology.

After successful technical development and validation, we have created a fully integrated product function that has received very positive responses from our customers. Expected results and effects:

The project results have taken the company a big step forward through a markedly improved sales offer to our customers where we can provide a complete solution for analysis of lameness in horses via a smartphone. The new product function has shortly after launch become very popular with our initial test group. Today, all our costomers use it and it generates more recordings than the original "straight line product function".

Approach and implementation:

Collection of training and reference data went according to plan. After building the database, algorithms and neural networks were developed to be able to create a robust analysis of movement asymmetry during lungeing. This very technically challenging part of the project resulted in high quality results that underwent validation and continuous benchmarking during the project.

Development of new signal processing algorithms gave new insights to how we can improve the product. User interfaces were developed iteratively via user feedback.