MELD is a use case project deployed as part of the IoF2020 project https://www.iof2020.eu
Early lameness is a considerable problem in the dairy industry. It causes pain and discomfort for the cow, while lowering fertility and milk yield for the farmer. Since current solutions come with high-initial costs and complex equipment, this use case utilises leg mounted sensors – measuring step count, lying time and swaps per hour – in combination with machine learning algorithms to identify lame cattle at an early stage. These data are analysed in the cloud and anomalies are sent to farmers’ mobile device to treat affected animals immediately and avoid further effects. As opposed to a general approach, this use case customises the data models to dynamically adjust as weather and farm conditions change. By detecting early lameness before it can be visually captured, treatment costs are decreased while animal welfare is improved.
MELD builds upon an existing trial for early lameness detection on a farm in South East Ireland and extend and integrate this deployment into the IoF2020 Trials. The current trial deployed on a farm with 150 cattle utilises leg mounted sensors and uses Machine Learning early lameness detection. MELD will integrate with the IoF2020 reference architecture and has identified the interoperability points to achieve this. Through this integration MELD will seek to make Ireland’s trial data available for analysis through other IoF2020 deployed algorithms and also utilise data from the currently deployed Herdsman+ trial in the UK to validate our Machine Learning approach to early lameness detection. MELD will also expand the Ireland testbed into three further testbeds in Portugal, Israel and South Africa.
The project is deploying equipment across 8 farms in Ireland, Portugal, Israel and South Africa and collecting data from over 900 cattle. This data is analysed through machine learning techniques to detect lameness in cattle at an early stage.
MELD has the following goals:
- Validate the solution on multiple vendor platforms and different environments;
- Integrate lame detection as a service into IoF2020 reference architecture and incorporate data from the dairy use case Herdsman;
- Scale the current solution to multiple sites across four countries with 1900 cattle;
- Develop machine learning algorithms for monitoring and early detection of anomalies;
- Combine state of the art technology with unique benefits to enhance the detection process, along with the cows’ welfare status;
- Ensure that the data collected is accessible only by the specific end-user to whom it is related
Expected results are:
- Required treatment time : -15%
- Reduced animal mortality
- Detection accuracy : 87%
- Days before visual detection : 3
- Increased productivity
- Improved animal welfare and reproduction efficiency
- Reduced usage of antibiotics