The nesteye platform
One edge system. A growing model library.
Cameras, sensors, local AI, and barn-level intervention work together today, while the same installed frame creates a path to new monitoring models.
Architecture
Built around the barn, not the cloud.
The system makes time-sensitive decisions locally and sends lightweight operational data to the dashboard.
Local operating loop
Decisions stay inside the barn.
Only lightweight status, trends, and alerts need to leave.
/01
Overhead IR cameras
PoE vision captures continuous flock behavior across the barn.
/02
Environmental sensors
Temperature, humidity, and air-quality readings add context to flock activity.
/03
Edge SoC box
On-device inference runs in under 500ms so time-sensitive work stays local.
/04
Light + audio intervention
Addressable LEDs and audio emitters disperse a confirmed pile with adaptive patterns.
Operator view
Five capabilities. One place to look.
Explore five current capabilities: piling, weight, environment, the integrator dashboard, and disease monitoring that surfaces mortality and gait signals for operator review.
/01 , Piling
Detect and disperse dangerous piling.
Overhead vision identifies developing clusters while adaptive light and sound can intervene inside the barn.
Repeated-frame confirmation helps separate a real event from ordinary flock movement before the system responds.
Confirmed event
Piling detected
Current capabilities
Clear about what is ready.
The current nesteye system combines the five capabilities below, including disease monitoring, mortality, and gait signals. Commercial onboarding remains pilot-led and barn-specific.
/01
Piling
Repeated-frame confirmation helps separate a real event from ordinary flock movement before the system responds.
/02
Weight
A broader overhead view reduces reliance on stressful catches and small, biased walk-over samples.
/03
Environment
Environmental signals stay connected to flock activity instead of living in a separate operating view.
/04
Dashboard
Only lightweight operational data leaves the barn; time-sensitive inference stays on the edge.
/05
Disease monitoring
Disease signals appear in the dashboard with barn context so teams can investigate and respond sooner.
Salmonella model
Flag fecal changes. Focus attention sooner.
The feces-based Salmonella model works on the existing frame. It needs no new hardware installation and provides an operator-review signal, not a diagnostic claim.
Research boundary
Fecal patterns can help prioritize attention.
Nesteye's feces-based Salmonella model flags visual changes for operator review on the existing camera and edge system. Supporting 2024 research found broiler activity can carry an early infection signal.
The model works as an operator-review signal. It is not a laboratory replacement and does not establish a diagnostic claim.
Read the research abstractDeployment
Start with one measurable pilot.
Every installation begins with barn fit, operating goals, and a shared definition of success.
Survey
Review the barn layout, camera coverage, network access, power, and operating goals.
Install
Mount cameras and sensors, connect the Edge SoC box, and calibrate the system for the house.
Pilot
Measure performance against an agreed operating baseline before broader deployment.
A typical nesteye system combines overhead PoE IR cameras, an environmental sensor array, an Edge SoC box, and addressable light and audio hardware for intervention.