# Can a camera cut the €8,570 lameness bill on Irish farms?

*Published 2026-10-01 by FarmAI Ireland*

Maybe. Lameness costs a 100-cow Irish herd €8,570 a year at moderate prevalence, and a camera over the parlour exit scores every cow's walk as she leaves. The one independent trial found it agreed with a human scorer 86.9% of the time, and caught about half the foot lesions — more than the human did. It is dairy-only, and no price list is published.

Autumn is when this matters most. Teagasc's advisory puts it plainly: risk peaks now because "cows are heavier in-calf and ground conditions become soft." Roadways that were fine in June start doing damage in October.

## What does lameness actually cost an Irish dairy farm?

This is the number the whole argument turns on. Teagasc published it with UCD on 30 April 2024. Cost per case, once treatment, lost milk and culling are counted:

| Condition | Cost per case |
|---|---|
| Digital dermatitis | €201.14 |
| White line disease | €226.67 |
| Foot rot | €308.14 |
| Sole ulcer | €536.52 |

Scale that to a 100-cow herd for a year and the range is wide. Low prevalence works out at €4,821.65. Moderate prevalence costs €8,570.26. Where it has got away on you, it is €14,191.40.

A 2022 Moorepark study put average prevalence on Irish dairy farms at 9%. The best-run farms in a 2021 study were under 5%.

So call it €8,500 a year on a hundred cows. That is the bill a lameness detection system is being asked to chip away at. Hold the figure in your head when a quote lands.

## How does camera-based lameness detection work?

The hardware is deliberately dull. One camera, mounted above the race where cows leave the parlour, pointing down. No collars, no tags, no leg sensors.

The software does the work. It tracks points on the cow's back and legs frame by frame and builds a picture of how she walks. It then compares that to her own normal and to the rest of the herd.

A mobility scorer stands at the collecting yard with a clipboard for an hour once a month. The camera scores every cow at every milking.

That frequency is the genuine advantage. Lameness does not arrive overnight — it builds. A monthly scoring round can miss a fortnight of a cow going downhill. A system watching twice a day cannot.

[CattleEye](https://www.cattleeye.com), the system with the most published validation behind it, was founded in Belfast and bought by the German dairy equipment firm GEA in 2024. It reads body condition score off the same footage, which is a second job for the same camera.

## What did the independent trial find?

This is the part vendors do not lead with, and it is the reason this article exists.

Researchers at the University of Liverpool ran a blind validation of the CattleEye system on three commercial dairy farms in northwest England and north Wales. It ran from November 2020 to February 2021 and was published in *Frontiers in Veterinary Science* in 2023. They compared the system against trained human scorers across 6,040 records.

The headline agreement was 86.9%. But the two statistics used to test whether that agreement is real rather than luck disagreed with each other. Cohen's kappa came out at 0.404, which reads as fair. Gwet's AC came out at 0.835, which reads as almost perfect. When the same data gives two answers that far apart, the honest reading is that the system is useful and not authoritative.

On detecting actual foot lesions — the thing really wrong with the cow — it had a sensitivity of 0.52 and a specificity of 0.81. In plain terms, it caught about half the lesions and rarely raised a false alarm. The human assessor it was measured against caught 0.29.

That comparison is the finding worth carrying home. The camera is nowhere near perfect. It beat a person doing weekly scoring anyway, because a person gets a few seconds per cow once a week and gets tired. A camera does not.

The authors flagged their own limits. The lesion dataset covered only 84 cows, and there was a single experienced assessor on one farm. It is one trial, on three farms, in Britain. We could not find an equivalent published Irish trial.

## Does this suit an Irish grazing herd?

Two honest caveats before anyone takes a quote.

The system needs a milking parlour with an exit race the cows walk through in single file. If you are running sucklers or drystock, this is not for you. Our piece on [what AI can and can't do for a suckler farm](/read/what-ai-can-and-cant-do-for-a-50-cow-suckler-farm) covers what is actually available at that end.

The second caveat is the validation farms were housed British herds. Irish cows walk roadways to grass for most of the year. Those roadways are both the main source of the lameness you are trying to catch and a different pattern of movement to what the model was tested on. Nobody has published what that does to accuracy here. Ask the vendor directly.

There is Irish research on the neighbouring approach. A UCD study at Lyons Farm, published in the *Irish Veterinary Journal* in 2021, used neck collar data from 164 Holstein-Friesian cows to predict lameness. The best model reached an area under the curve of 85%, with sensitivity and specificity of 78%. The authors cautioned that the behavioural differences between sound and unsound cows were small. They also said the models still need testing in commercial pasture-based herds.

If you already run collars, that capability may be sitting in the system you own. Our [SenseHub review](/read/sensehub-ireland-full-review) and the [Halter piece](/read/does-halter-actually-work-irish-dairy-farmers) both cover what those platforms flag.

## What does it cost, and what should you ask?

Nobody selling lameness detection publishes a price list, so we are not going to invent one. CattleEye's own site says installation is "up & running in a day" with "no specialist engineers to wait for", and points farmers at a savings calculator rather than a figure. Pricing is by quote. It will depend on herd size and on what you already have installed.

Which means the work is in the questions. Before you sign anything, ask:

1. **What is the total first-year cost** — camera, installation, software subscription and any annual licence — as one number?
2. **Is the subscription per cow or per farm**, and what happens to it if the herd grows?
3. **Does it need a hardwired connection at the parlour**, or will farm wifi carry it?
4. **Is there a named Irish reference farm** you can ring? Not a case study — a phone number.
5. **What has it been validated against on grazing herds?** If the answer is the Liverpool study, you now know exactly what that study found.
6. **What is the exit clause** if it does not earn its keep in year one?

Then run the answer against the €8,570 figure. If the system costs a modest share of that and cuts your prevalence by a couple of points, it stands up. If it costs most of it, it does not.

## Frequently asked questions

**Does camera lameness detection work on a grazing herd?**

Nobody has published an Irish trial on it. The one independent validation, from the University of Liverpool in 2023, ran on three housed British herds. The system needs a parlour exit race, so it is dairy-only. Ask any vendor what they have validated on pasture-based cows before you commit.

**What does lameness cost per cow in Ireland?**

Teagasc and UCD costed it in April 2024 at €201.14 a case for digital dermatitis, up to €536.52 for a sole ulcer. For a 100-cow herd across a year that is €4,821.65 at low prevalence and €14,191.40 at high prevalence. Average Irish prevalence was 9% in a 2022 Moorepark study.

**Can I get TAMS support for this?**

Check directly with DAFM or your Teagasc advisor before assuming either way. TAMS investment lists change between tranches, and eligibility for a software subscription is different to eligibility for hardware. Do not order on the strength of a salesperson saying it qualifies.

## Where to get help

Animal Health Ireland runs the [Hoof HealthCheck programme](https://teagasc.ie/news--events/daily/new-programme-to-tackle-lameness-in-dairy-herds/) in partnership with Teagasc and UCD. It works with ten pilot dairy farms around the country, each getting tailored vet, hoof trimmer and advisory support. AHI programme manager Dr Michelle McGrath puts it straight: "Lameness is a serious concern and should be treated as such."

Talk to your Teagasc advisor and your own vet before you spend anything. No camera fixes a roadway. Teagasc's autumn advisory lists the cheap wins first — standing water, broken surfaces, sharp turns, and grass build-up blocking drainage on verges. Every one of those costs less than a subscription.

## The bottom line

If you are milking through a parlour, and you treat a flag as a prompt to lift a foot rather than a diagnosis, a camera will find cows earlier than you will. If you are running sucklers, or your roadways are the actual problem, spend the money on the roadways first.

---

## Sources

- [Teagasc — Shedding light on the cost of lameness](https://teagasc.ie/news--events/daily/shedding-light-on-the-cost-of-lameness/) — Teagasc and UCD costings for each lameness condition and for the 100-cow herd, published 30 April 2024
- [Teagasc — Lameness on the dairy farm: autumn is peak risk period](https://teagasc.ie/publications/lameness-on-the-dairy-farm-autumn-is-peak-risk-period-php/) — Teagasc advisory on why autumn carries the highest lameness risk on Irish farms
- [Teagasc — New programme to tackle lameness in dairy herds](https://teagasc.ie/news--events/daily/new-programme-to-tackle-lameness-in-dairy-herds/) — Animal Health Ireland's Hoof HealthCheck programme, run with Teagasc and UCD
- [Frontiers in Veterinary Science — Initial validation of an intelligent video surveillance system for automatic detection of dairy cattle lameness](https://pmc.ncbi.nlm.nih.gov/articles/PMC10299827/) — University of Liverpool independent validation of a camera mobility-scoring system, 2023
- [Irish Veterinary Journal — Prediction of lameness using automatically recorded activity, behaviour and production data in post-parturient Irish dairy cows](https://pmc.ncbi.nlm.nih.gov/articles/PMC7868012/) — UCD Lyons Farm study of collar-based lameness prediction in 164 Irish dairy cows, 2021

---

*Source: [FarmAI Ireland](https://farmai.ie/read/camera-lameness-detection-irish-dairy-cost)*