the Danish Horse Levy Fund - Audio data for early detection of lameness in horses

Principal Investigator: Franziska Hakansson, Danish Technological Institute 

Project Participants

The project is led by Senior Consultant Franziska Hakansson, who has a strong interdisciplinary background in ethology, animal welfare, and applied machine learning. The project is conducted in collaboration with veterinarians from the University Hospital for Large Animals (University of Copenhagen), who contribute with expertise in equine pain and orthopaedics and facilitate data collection.

Purpose

The overall aim of this pilot project is to investigate the value of audio data from a horse's gait for the early and objective identification of lameness. The project will explore whether machine learning models can be trained to distinguish between sound and lame horses based on audio recordings, with the long-term goal of developing user-friendly tools that can promote health, welfare, and timely treatment for horses.

Method

In the project, a dataset of audio recordings will be collected from horses undergoing lameness examinations at the University Hospital for Large Animals. Data will be analysed using machine learning to identify the patterns and characteristics in the sound of the horse's gait that are associated with lameness. The goal is to develop a model that can, with high accuracy, distinguish between audio recordings from lame and sound horses.

Timeline and Publication

The project will run during 2026, with data collection in the first half and data analysis/model development in the second half of the year. Results are expected to be published in early 2027 on the Danish Technological Institute's website via a project page. The results from the project will be made available free of charge to all businesses and stakeholders in the Danish equine sector.

Funding

The project has received a grant from Hesteafgiftsfonden (the Danish Horse Levy Fund).

Læs om projekets beskrivelse på dansk her.