The Tricky Business of Dairy Forecasting


Photo caption:Pictured at the launch of the Smart Appi Project are Enda McDonnell, Enterprise Ireland; John Fitzgerald, Glanbia; Minister Richard Bruton, TD; Eric Robson, Waterford Institute of Technology; Liam O’Flaherty, Dairygold; Brian Enright, Irish Cattle Breeding Federation; and Dr Michael D Murphy, Cork Institute of Technology.

Cows in pastures munching on grass are a common sight in Ireland, but rare worldwide where many cows spend their lives indoors or in enclosures eating crops such as maize. Since Irish dairy farming is unusual, the new Smart Appi Project is investigating ways to track and accurately predict the volume and quality of milk that is produced.

Dr Michael D Murphy from the School of Mechanical, Electrical and Process Engineering in Cork Institute of Technology (CIT) is working on mathematical models as part of this project. Farmers and producers find it difficult to accurately predict how much milk will be produced on a weekly or monthly basis, according to the engineer.

Unreliable supply

Ireland and New Zealand are the only two countries whose dairy industry is purely pasture-based. Being exposed to the weather and relying on the growth of grass means that the volume and quality of milk varies.

“There are prediction methods, but they have been used for the last couple of decades and wouldn’t be terribly advanced,” explains Murphy.

The Smart Appi Project has teams from across Ireland working on ways to improve data analysis and modelling systems to allow farmers and processers to forecast milk production.

Investing in software

CIT is working on the project with the Telecommunication Software and Systems Group (TSSG) in Waterford Institute of Technology (WIT), Teagasc, Glanbia and Dairygold. Enterprise Ireland has invested almost €445,000 to develop the new software.

The project teams are already working on the project. CIT is building on its previous experience of modelling milk yields on farms by analysing equipment such as milking machines and cooling tanks.

“Several models would already pull information from databases and predict milk yields for processers or at a farm level,” according to Murphy.

TSSG will be organising the data, which is currently spread out in multiple databases.

Planning ahead

The project will develop software that can be used by processers and dairy farmers to plan ahead. Murphy explains this will not only help large processors such as Glanbia and Dairygold, but also farmers who “get paid per litre of milk produced so will be able to forecast how much money they are going to make”.                                                       

National milk production quotas will end in April 2015. Farmers will be able to increase their milk production, which could the price of milk could drop. The project teams feel that being able to reliably forecast variations in milk supply could improve competiveness and boost employment in the sector.