Computer vision for farms, glasshouses, and growers.
The herd is in the field, the crop is in the polytunnel, the grain is on the belt. You already walk it, count it, check it by eye. I build vision systems that do the same job, continuously, from cameras on the gate, the gantry, or the drone. Built on your hardware, code in your repo.
Or copy: [email protected]
What's actually happening on your farm right now.
You're counting cattle from the pickup. You're walking glasshouses with a clipboard. You're sampling pallets of fruit at intake and hoping the sample is representative. The work gets done, mostly. The bits that slip through cost you.
A sick animal spotted three days late. A pest outbreak found when it's already in the next bay. A bad batch of fruit that made it to the customer because nobody saw the bruising at intake. The cameras to catch all of this exist, are cheap, and would pay for themselves inside a season. They just need software that knows what to look for.
What I build for farms and growers.
Livestock monitoring
Cameras in the shed, the yard, or on the gate count animals, flag distress behaviour, watch for calving or lambing, and log feeding patterns. Works for cattle, pigs, sheep, and poultry. Reports drop into an email or your existing farm software.
Quoted on scope.
Fixed price in writing before any work starts. Hardware, on-site installation, and any opted-in subscriptions quoted separately at cost.
Crop counting and yield estimation
Camera mounted on a gantry, a quad bike, or a tripod. Counts fruit on the trees or vines, estimates yield per row, flags rows that are underperforming. Works for soft fruit, top fruit, citrus, olives, vines.
Quoted on scope.
Fixed price in writing before any work starts. Hardware, on-site installation, and any opted-in subscriptions quoted separately at cost.
Glasshouse and polytunnel monitoring
Cameras at row ends or on rails flag pest infestations, disease symptoms, and irrigation issues. Daily summary by email. Particularly useful for high-value crops where a missed outbreak costs a whole bay.
Quoted on scope.
Fixed price in writing before any work starts. Hardware, on-site installation, and any opted-in subscriptions quoted separately at cost.
Quality at intake
Camera over the grading belt or the pallet line photographs and assesses every batch coming in. Flags damage, foreign objects, off-spec produce. Searchable log when a customer raises a complaint two weeks later.
Quoted on scope.
Fixed price in writing before any work starts. Hardware, on-site installation, and any opted-in subscriptions quoted separately at cost.
A few of the situations this system handles.
A farm with cattle in winter housing where overnight calving needs attending in time. The system watches the shed, recognises calving behaviour, and alerts the farmer's phone. Yard headcount stays accurate without manual counts.
A polytunnel or glasshouse growing a high-value crop. The system watches the rows for pest and disease symptoms and irrigation issues. The grower gets a daily summary email with rows that need attention.
A grading line at intake for soft fruit, top fruit, or root vegetables. The system reads the line, flags damaged or off-spec produce, and logs every batch with images for the customer-complaint paper trail.
A mixed enterprise with livestock, glasshouse, and intake operations on one site. Each setup runs locally, with the farmer getting one consolidated morning email covering whatever happened in the last 24 hours.
Why this, why me.
Built for the kit you have.
Most farm CCTV from the last few years works for this. New camera installs are simple and cheap where needed. The clip review tells you what fits your setup.
Works without internet on the ground.
The system runs locally. Alerts go out over 4G if the broadband drops, which it will. Designed for rural sites where connectivity is occasional.
The code is yours.
No subscription. No farm-tech vendor lock-in. You own what you paid for.
Built for one farmer or a small team.
No dashboards that need a degree to read. Most outputs are a phone notification or a one-page email.
Fixed price, in writing.
A call, then a build.
We talk
Fifteen minutes. You tell me what you farm, what cameras you have, and what you'd most like to know that you currently don't.
I build
Two to six weeks for most farm installs. Livestock work tends to be faster; crop work depends on whether we're working with an existing camera setup or specifying new ones.
Every quote breaks out the build fee, hardware, install labour, and any subscriptions in writing before you commit. One total, every line visible.
The things people ask before we start.
Will this work without good broadband?
Yes. The system runs on the farm. Alerts go out over 4G if needed, and the system stores its own data locally rather than relying on the cloud. Internet is required only for software updates, which happen monthly at most.
What about night work?
Most modern cameras have infrared. We confirm during scoping whether yours work in the dark. If new cameras are needed for night-time monitoring of housed livestock or perimeter work, infrared-capable units are inexpensive.
Can it identify individual animals?
With ear tags visible, yes. Without ear tags, the system tracks animals in a group but doesn't name them. Many of the most useful jobs (headcount, calving or lambing behaviour, feed pattern monitoring) don't need individual identification.
Does it work for non-UK farms, including Morocco and Europe?
Yes. The kit and the software are geography-neutral. Prices quotable in GBP, EUR, or MAD on request. The use cases are often stronger outside the UK, especially for olive, citrus, and stone fruit operations.
Does it integrate with my farm management software?
Often yes. Most modern farm software has an API or accepts an import. We discuss your specific platform on the call. If you're not using anything yet, the system can deliver everything as plain email reports.
What happens if the system flags something that turns out to be nothing?
Every alert is a still or a clip you review before acting. The system suggests, you decide. False positives get fed back into the system to reduce them over time. The early weeks always have more noise than the later ones.
How quickly can it be running on my farm?
Two to six weeks for a single farm. Livestock work tends to be faster, usually two to four weeks. Crop work depends on whether we're using existing cameras or specifying new ones for the row ends or gantries, and can run four to six weeks. Multi-farm setups for the same enterprise run faster per site after the first.
Tell me about your farm.
Email me a sentence about what you farm and what you'd most like the cameras to tell you. I reply the same day with a rough scope and a fixed price in writing within a few days of a call.
Or copy: [email protected]