Logo Lanfrica
  • Home
  • Atlas
  • Insights
  • Docs
  • Sign in

© 2026 Lanfrica. All rights reserved. All copyrights of the resources shown on the Lanfrica website belong to the original copyright holders, unless explicitly stated otherwise.

sellomakgatho121/Agentic-Poultry-Monitoring

Domain:

agriculture

Record type:

softwareproject
Creator:
sel
Host:
Three-layer PLF engine monitoring 1,242 laying hens at Boonducks Farm, South Africa. YOLOv8-nano vision, XGBoost acoustic stress detection, TimescaleDB telemetry — all on Jetson Orin Nano edge AI. # Agentic Poultry Monitoring — Boonducks Farm PLF Engine **A three-layer Precision Livestock Farming (PLF) engine deployed at Boonducks Farm, South Africa.** Monitors 1,242 laying hens across 8 coops using edge AI computer vision, acoustic stress detection, and environmental IoT telemetry. | Coop Type | Count | Capacity | |-----------|-------|----------| | 🏭 Cage Coops | 2 (C1, C2) | 312 + 310 hens | | 🌿 Deep Litter Coops | 6 (L3–L8) | 104 hens each | | **Total** | **8 coops** | **1,242 hens** | --- ## Architecture ### Layer 1 — Edge (In-Coop Inference) - **NVIDIA Jetson Orin Nano Super** (67 TOPS, TensorRT-optimized) - **YOLOv8-nano** bird detection + Huddling Index from overhead cameras - **XGBoost** acoustic stress classification from USB/I2S microphones - Only anomaly JSON forwarded to fog layer — preserves bandwidth ### Layer 2 — Fog (Farm Office Server) - **TimescaleDB** hypertables for environmental telemetry - **Express.js** API serving continuous aggregates to dashboard - **Modbus RTU RS485** daisy-chain for XY-MD02 temp/humidity and NH3 probes - **Hikvision 4MP PoE** cameras for vision feed ### Layer 3 — Cloud (Future) - Federated learning across farms - Remote dashboard access - Automated model retraining from cross-farm outbreak patterns --- ## Repository Structure ``` ├── vision_analyzer.py # YOLOv8-nano bird detection + Huddling Index ├── acoustic_analyzer.py # XGBoost stress classification (MFCC features) ├── telemetry_worker.py # Modbus RTU sensor polling + TimescaleDB ingestion ├── server.js # Express.js API server (sim/live modes) ├── schema.sql # TimescaleDB hypertable definitions ├── client/ # Dashboard frontend (React + Vite) ├── scripts/ # Deployment and utility scripts ├── Dockerfile # Containerized deployment ├── docker-compose.yml # Multi-service orchestration ├── SETUP_GUIDE.md # Complete hardware + software installation guide └── …

Visit

github.com

Tasks

image classificationcomputer vision

Tags

agriculturecomputer-visionedge-aiiotjetson-orinmodbuspoultryprecision-livestock-farmingsouth-africatimescaledb+2

Similar

Development of an Agentic AI Framework for Power Infrastructure Monitoring and ResiliencyLeoKips/Agentic-AIago60hz/agentic-capital-velocityCharlesMfouapon/agentic-liquidity-routerM0iem/agentic-rag-tourism-moroccojohncharlessarwi-dot/agentic-tourism-platform

Development of an Agentic AI Framework for Power Infrastructure Monitoring and Resiliency

Monitoring critical infrastructure (CI), particularly national power distribution networks in develo

LeoKips/Agentic-AI

This project entails developing a Search Agent that analyzes the most recent policies for various in

ago60hz/agentic-capital-velocity

An open-source agentic orchestration layer for African B2B payments using ERC-8004 identity framewor

CharlesMfouapon/agentic-liquidity-router

Autonomous AI agent for cross-border payment routing across African mobile money rails and onchain l

M0iem/agentic-rag-tourism-morocco

# Agentic RAG Tourisme Maroc Projet de fin de module: assistant intelligent de tourisme au Maroc ba

johncharlessarwi-dot/agentic-tourism-platform

Complete Agentic Tourism Platform - AI Agents for SEO, Leads, Marketing & Operations in East Africa