George McMaster began his farming career milking eight cows in a byre, relying on his eyes to spot anything wrong with the cattle.
He still watches his herd — now 300-strong — but through a feed from an AI camera in the milking parlour at his farm near Islandmagee in County Antrim.
For a third-generation farmer, he told BBC News NI, it is “absolutely incredible to move from all hand-done to sitting on the phone, watching what’s happening outside”.
What the camera actually does
The device runs an algorithm that assigns every cow her own ID, records a short video of the animal each day, analyses it, and flags any changes — often before they are visible to the human eye.
His son Jack, who is taking over the farm, got involved with the CattleEye app when a local developer approached him. His account of what it found is the part worth quoting.
“There’s cows there we would never have thought of as lame when we were just watching them every day, but whenever the camera’s watching they’re picking them up so much earlier,” he said — in time “to check cows before they lose condition and get sick”.
And the economics follow the welfare: “if cows are in better health, we get more milk and they’re happier — happy cows make happy farmers.”
Note what the machine is beating here. Not an inattentive farmer: a third-generation one, looking at the same animals every single day. Lameness develops gradually, and gradual change is exactly what human observation is worst at, because each day looks like the last.
The thing that had to come first
Caleb Howard manages 90 Jersey cows near Coagh in County Tyrone. They filter in and out of a robotic milking machine when they choose to, rather than being herded in at set times. Like the McMasters, he thinks “everything is going down the robotic route” for efficiency.
But he is blunt about the precondition.
“Ten years ago, in our area, phone signal was rubbish, the broadband wasn’t great, it was just like a slightly faster dial-up,” he said. “So now everything is much simpler and much easier-used when you’ve got a good internet connection.”
Getting high-speed broadband was what let him open a farm shop with an electronic payment system. The connection came through Project Stratum, delivered by Fibrus. Colin Hutchinson of Fibrus called the economic impact “massive”: “Being online and having good broadband access is really crucial now. And it’s enabling businesses like this to thrive in really rural locations.”
“Facebook for cows”
Robert Smyth runs 100 dairy cows near Randalstown and has used the FarmDrive app for four years.
“Beforehand you’re writing things down, bits of paperwork in the cattle house and writing them down in the office, whereas now you just do it at that time,” he said. “And then whenever you have inspections coming up, all you have to do is go into your compliance section, select reports, and there’s your reports for your medicine, for your meal records, for everything.”
Developer Gareth McDonald, who comes from a farming background and studied computer science, describes FarmDrive as “Facebook for cows” — each animal gets a profile with her records in one place and a photo.
He is firm about the boundary. “It’s allowing us to work a lot more quickly and efficiently, and on the farming side, it’s basically allowing the farmer to see the right information at the right time, where it’s kind of that adviser to the farmer. We’re very much of the opinion we don’t want the AI making the decision for the farmer.”
Smyth agrees on where the value sits: “The AI really should be looking towards how to manage your farm better and how to reduce your inputs and increase your outputs.”
The long view
Almost a century ago the County Down-born inventor Harry Ferguson transformed farming with the three-point linkage still used on tractors today. Ian Marshall, former president of the Ulster Farmers’ Union, thinks AI is similarly revolutionary — and frames it around data rather than machinery.
“When we take all the datasets, then I think we can deliver on the three pillars of sustainability — economic, environmental and social,” he said.
What this means for Bangladeshi farming
The sequence in this story is the useful part, and it runs in the opposite order to how these things are usually sold here.
Nobody bought AI first. The connection came first — through a public programme — and the applications arrived afterwards, built by local developers who approached individual farmers. Howard could not have run a robotic parlour or an electronic till on what he had before, no matter how good the software was.
Bangladesh has a dairy sector with the same problems these farms had: gradual illness missed, records on paper, compliance evidence assembled in a panic before an inspection. The tooling is not exotic — a camera, an app, a profile per animal. The question is whether the rural connection exists to carry it, and that is decided well before any farmer chooses a product.
