Livestock Farming & AI Glossary
The terms that decide whether a batch makes money, defined in plain language — the same way they're used across DeepLoop.
FCR (Feed Conversion Ratio)
Kilograms of feed needed per kilogram of weight gained. The single number a pig finisher or broiler batch is judged on — lower is better, and small drifts compound into real money.
ADG (Average Daily Gain)
How much weight an animal adds per day. Falling ADG is often the first quiet sign that health, feed or climate has gone wrong.
Livability
The share of a flock or batch that survives to market. The headline number for a broiler house; every early detection defends it.
Lay rate
Eggs produced per hen per day, usually as a percentage. Layer flocks are managed to hold the lay-rate curve at peak as long as possible.
Floor eggs
Eggs laid outside the nest. A rising floor-egg count often travels with stress or feather-pecking, and each one costs labour and downgrades.
Feather pecking
Birds pecking each other's feathers — behaviour that spreads through a layer flock if not caught at the first pecks, dragging lay rate and welfare with it.
Days to market
How long a batch needs to reach target weight. Every sickness event or feed problem quietly adds days — and days are money in housing, feed and planning.
Condemnations
Carcasses or parts rejected at the processing plant. Footpad lesions, hock burn and disease show up here weeks after the barn problem that caused them.
Mortality record
The running count of dead and culled animals per house. Feeds livability, FCR and settlement — and is only useful when it's current and correct.
Computer vision
Software that reads camera images — counting animals through pile-ups, estimating weight, tracking gait and behaviour. The sensing layer of AI livestock monitoring.
Precision livestock farming (PLF)
Using continuous per-animal measurement — cameras, sensors, sound — to manage livestock on data rather than averages. AI monitoring is PLF's current front edge.
Closed-loop autonomy
A system that doesn't stop at alerting: every detection updates records, recalculates targets and adjusts feed or climate — watch, understand, decide, act, learn. The principle DeepLoop is built on.