v1.0 Design Guide · Smart Agriculture

Smart Agriculture Environmental Monitoring Design Guide

A comprehensive engineering reference for procurement, implementation, delivery, and operations of smart agriculture environmental monitoring systems — covering greenhouses, open fields, orchards, pastures, and aquaculture.

12
Chapters
8+
Application Scenarios
5
Interactive Calculators
10,000+
Max Monitoring Points

System Overview

The Smart Agriculture Environmental Monitoring System is a comprehensive engineering platform designed for facilities agriculture (greenhouses, tunnels), open-field cultivation, orchards and tea gardens, pastures, and aquaculture operations. It performs continuous monitoring and data governance over weather, soil, crop microclimate, and facility/equipment state, forming a closed loop of multi-point sensing → edge control → platform analytics → alarm linkage → farming operations feedback.

This guide is intended for EPC contractors, system integrators, and farm-owner self-build projects. It covers the complete engineering lifecycle from site survey and architecture design through sensor selection, installation, commissioning, and long-term operations and maintenance. The system is designed to operate across a wide environmental envelope: ambient temperatures from -20°C to +55°C, relative humidity from 10% to 100%, wind speeds up to 35 m/s, and full exposure to heavy rainfall and storm conditions.

Core Value: The system's "engine" is not just sensors — it is representative placement, calibration and maintenance discipline, data validity labeling, and operable alarms that reduce false actions and enable trustworthy decisions that drive measurable improvements in resource use and crop outcomes.

Positioning & Goals

The system provides a reliable and maintainable sensing-and-control foundation for precision irrigation and fertigation, disease and pest risk early warning, extreme weather response, and yield and quality improvement. By delivering actionable alarms, validated data, and farm-task closure (who did what, when, with evidence), it reduces resource waste and operating cost while improving traceability and compliance.

Applicability & Boundaries

System Architecture

The overall system architecture follows a layered design from the physical sensing environment through edge processing, communications, platform services, and application integration. Each layer has clearly defined responsibilities, interfaces, and quality boundaries. The architecture supports deployments ranging from small single-farm installations (30–100 monitoring points) to large multi-farm, multi-tenant operations with 1,000–10,000+ points.

System Architecture Layered Diagram

Figure 0.1: Smart Agriculture Environmental Monitoring System — Overall Architecture (7-Layer Diagram)

The architecture distinguishes between core components (sensing, edge, platform ingestion, QC flags, alarms, dashboards), optional enhancements (AI pest models, satellite backup, video analytics, advanced optimization), and supporting dependencies (network, UPS, power distribution, lightning grounding, cabinet infrastructure, fire linkage, physical security). This separation allows projects to start with the core and add capabilities incrementally without redesigning the foundation.

Key Data and Control Flows

Main Functions

The platform delivers nine primary functional capabilities, each designed to convert raw sensor data into actionable farm operations. The diagram below presents these functions in a nine-grid overview, showing the input, processing logic, and output for each capability.

Main Functions Overview 9-Grid

Figure 0.2: Main System Functions Overview — Nine-Grid Capability Map

FunctionCore ValueKey ImplementationAcceptance Focus
Multi-protocol Device AccessReduces vendor lock-in; accelerates rolloutMQTT/HTTP/CoAP/LoRaWAN NS, Modbus RTU/TCP; unified device modelOnboarding time, telemetry completeness, retry behavior
Data Governance & Quality FlagsPrevents wrong irrigation/fertilization decisionsRange checks, ROC checks, redundancy cross-check, calibration trackingQC flag correctness on seeded fault cases; auditability
Edge Buffering & Offline ContinuityAvoids data loss during outagesLocal store-and-forward; local rule fallback; time sync strategyOutage simulation with recovery; data gap limits
Alarm Center & Actionable NotificationsTurns data into operations; reduces alarm fatigueRule templates by scenario; severity, dedup, escalation, suppression windowsFalse alarm rate, MTTA, closure evidence
Closed-loop Environmental ControlStabilizes greenhouse climate; improves qualityEdge PID/logic constraints; safe mode; manual overrideResponse time, overshoot limits, safety interlocks
Irrigation/Fertigation Decision SupportSaves water/fertilizer; improves yieldSoil moisture + ET0 + crop stage + irrigation capacity modelMeasurable reduction, recommendation correctness
Extreme Weather Response PlaybooksReduces loss from frost, heat, wind, hailThreshold rules + forecast ingestion; SOP linkageDrill test; response timelines
O&M: Device Health & CalibrationKeeps data trustworthy long-termBattery health, comm quality, sensor drift detection, maintenance schedulesMTBF improvement, calibration compliance rate
Open APIs & IntegrationConnects to farm management, traceability, ERPREST/Webhook, MQTT topics, event schema, RBACAPI SLA, security controls, integration test cases

Chapter Navigation

This guide is organized into twelve chapters covering the complete engineering lifecycle. Use the cards below or the left sidebar to navigate to any chapter.

Typical Project Deliverables

A complete smart agriculture environmental monitoring project should produce the following engineering deliverables, which serve as both acceptance criteria and long-term O&M references.

Success Criteria: Reduced water and fertilizer usage (typically 5–20%), fewer crop losses due to extreme weather (measurable by incident reduction), and improved labor efficiency (reduced manual scouting). Budget sensitivity is high — prioritize lifecycle cost (LCC) and maintainability over premium hardware.