Poultry Spotlight: Monitoring Livestock Count, Behaviour and Health with AI

Can you distinguish one bird from a lot of 500? 5,000? 50,000?

What about monitoring their behaviour across their lifetimes?

Poultry monitoring is particularly tricky because of its high population density in confined areas as well as their similarity in appearance and behaviour, making it difficult to track behaviour, symptoms of health condition. When birds of differing ages are grouped together, it can make them more difficult to care for.

More farms have turned to AI surveillance set-ups to manage their poultry flocks. This typically involves cameras positioned at strategic locations throughout the farm can capture continuous visual data in poultry farms, and AI algorithms that process the visual data to identify and count individual birds.

The Difference Between Spot Checks and Always-On Surveillance

Farmers providing flock care are able to monitor the birds based on their routines. Vision-based AI systems are fill in the gaps by monitoring the flock 24/7. Additional capabilities such as identifying individual birds can help farmers give birds with abnormal behaviour more attention.

Counting: Coops that have defined flock sizes and age ranges enable farmers to keep track of flock numbers. Alternatively, with the use of camera systems with AI capabilities, farmers can routinely check flock size and obtain real-time numbers of birds with certain characteristics, such as illness symptoms and size.

Behaviour & Health: Timed checks on bird flock can identify injuries and obvious symptoms of illness, but subtler changes in behaviour over time can be overlooked. For example, reduced movement, aggression or feeding may indicate stress, illness, or overcrowding.

By the same token, by using cameras to monitor birds constantly, AI systems can identify early signs of disease outbreaks in coops by analysing heatmaps of flock activity or spotting lethargic or isolated birds.

3 Ways to Use Poultry Farm Surveillance Data

Video surveillance of poultry farms can capture bird count alongside other visual characteristics. The combination of this data can lead to actionable cues and insights. For example, being alerted that an individual bird is sick enables a farmer to give it the attention it needs; being alerted that a significant numbers of birds show abnormal symptoms could be a signal for farmers to investigate an environmental issue within the farm.

Below are some ways AI can help with monitoring poultry farms:

Count and Categorise Birds within a Flock

• Bird count according to physical characteristics like size and plumage colour
• Bird count according to movement patterns such as active versus resting birds, high versus low activity birds, or other behavioural tendencies
• Bird count according to physical anomalies or behavioural deviations

Assess Physical Appearance of Individual Birds

• Size and Weight: Detect changes in body size or growth rate by analysing individual bird dimensions over time
• Feather Condition: Identify feather loss, which could indicate stress, disease, or aggression
• Visible anomalies: Monitor wounds, swelling, deformities or discolouration, which may signal health issues like anaemia or illness

Monitor Behaviour and Health Indicators of Individual Birds

• Posture: Analyse standing, sitting, or lying positions to detect discomfort or difficulty in movement
• Movement and Activity Levels: Track birds' activity to identify lethargy, reduced movement, or abnormal pacing
• Feeding and Drinking Behaviour: Observe how often and how long birds spend at feeders and drinkers, detecting potential health issues
• Grouping and Isolation: Detect birds separating from the flock, often an early sign of sickness
• Aggression or Pecking: Monitor interactions to identify aggressive behaviours that could harm flock health
• Clustering or dispersion: Such behaviours might indicate poor environmental factors

Speak to an AI expert about livestock monitoring - book a meeting now.

Want to explore similar use cases? Tell us your requirements.