Innovative technology for better lameness detection in dairy cattle, enhancing animal welfare, strategic monitoring, and reducing chemical treatments.
The challenge
Lameness is recognised as the primary animal welfare issue in dairy farming. At any one time, around 30% of dairy cows across the UK are considered to have a degree of lameness1. This issue has a wide-ranging impact, including on the quality and yield of milk production, animal fertility, and their feed intake, lifespan and culling rates. Combined, this can result in a far higher environmental impact of milk production and an economic effect on farmers of around £300 per incidence of lameness.
Although it has multiple causes, the key factor for lameness in most herds is endemic diseases, such as digital dermatitis (DD), sole ulcer and white line disease. With early detection and prompt intervention critical to effectively controlling and treating lameness, current methods, such as mobility scoring, are often too reactive to an already established problem in the animal.
There’s also a need for farmers to more effectively monitor and reduce their use of chemical treatments, not only for economic reasons, but to avoid unnecessary contamination into local soil and waterways.
The Hoofcount solution
Hoofcount are already a well-established name in the UK dairy sector. Since 2012, the company has been a leading provider of automatic self-cleaning, emptying and preparing disinfection footbaths, used by farmers as a key tool to prevent disease in their animals.
As part of their efforts to improve prevention and detection of lameness in cows, as well as increase the efficiency of on-farm treatment processes and chemical use, Hoofcount developed a novel AI-led solution to more effectively identify and monitor early signs of DD and similar diseases. By using hi-tech cameras and computer modelling, Hoofcount’s system aims to bring new technology to the forefront of animal welfare management.
Project scope
Hoofcount secured funding in 2022, working alongside the UK Agri-Tech Centre and the Centre for Machine Vision at the University of the West of England. Across four demonstration farms, the project tested the use of a hoof-level camera at the exit of the footbath to capture multiple images of the underside of each hoof as an animal passes through.
The electronic identification (EID) system can successfully identify which animal is moving through the footbath, as well as which hoof is shown in each image, before ascertaining the existence of any very early-stage lesion within the hoof. This allows farmers to build a detailed picture of each animal’s ongoing health and adjust their treatment as necessary. A further benefit is that entirely healthy cows are not required to use the footbath, providing time, cost and chemical use savings to farmers with a more targeted process.
Beyond demonstration, the project undertook a range of system development and optimisation work. Primarily, this involved the creation of a library of thousands of images that inform the algorithm within the camera system for what is in each picture and where a potential lesion can be identified more accurately.
Further activity was completed to assess the impact of crucial factors such as lighting within the cowshed, the walking direction of animals and the setup of footbaths, and issues surrounding limited internet speed. The data gathered from this work proved valuable in establishing guidelines for use to provide the best outcomes for users.
The project also sought the input and feedback of leading experts from across the dairy sector. Anthony Marsh, Director at Hoofcount, commented:
“It was vital that we validated the potential and benefits of the system through specialists in cow health, to inform our work and help showcase the technology. We were pleased to be able to work alongside leading lameness vet, Nick Bell, as well as gaining the input of other vets, nutritionists and hoof trimmers. The feedback we received was extremely encouraging, particularly around the accuracy of the imaging system for identifying potential lesions at a far earlier stage than at present.“
“This input helped shape our development plans as we progressed, and, most importantly, gave us the backing for the impact the system could have on the dairy sector. This potential is exactly why the support from the Farming Innovation Programme was so valuable. In addition to the funding that has enabled the project to move forwards, we’ve also seen the benefits of working alongside their innovation experts for identifying opportunities for collaboration, patents and securing skilled hires.”
What’s next?
Following the completion of the project, Hoofcount are combining further research and development with an expanding commercial effort.
They are continuing to work with the demonstration farm sites to help gather more images to further boost the accuracy of the lesion identification system and continue to track the welfare of the animals on site. These sites will also be key for other technology development efforts in the future.
But, as Anthony added, there’s now a commercial opportunity for Hoofcount to focus on too:
“Having proven the effectiveness of the technology, we are now working with our existing footbath customers, as well as new connections, to add the camera system into their site. With international demand also increasing, particularly in the USA and Europe, we will continue our efforts to optimise the system and showcase the benefits to a wider audience.”
