IoT Connected Greenhouse - Smart Agriculture

Overview

Design and realization of a connected greenhouse using IoT technologies to optimize agricultural growing conditions while minimizing resource consumption. Complete system integrating environmental sensors, LoRaWAN transmission, cloud visualization, and actuator automation.

Context & Problem Statement

Environmental Challenges

Agriculture is on the front line of climate change, facing major challenges:

• Carbon impact: 10-12% of human-caused GHG emissions
• Water scarcity and increasingly harsh climate conditions
• Need to optimize resources (heating, irrigation) while maintaining productivity

Problem Statement

How to use IoT technologies to improve agricultural productivity in greenhouses while optimizing resource usage?

Objectives

Develop a functional prototype to:

• Monitor greenhouse environmental conditions in real-time
• Transmit data to a cloud platform for analysis
• Automate actuators (heating, ventilation, irrigation) based on actual needs

System Architecture

1. Environmental Sensors

BME680 (I2C):

• Temperature
• Air humidity
• Atmospheric pressure
• Air quality (IAQ)

LM94021 (ADC):

• Temperature (reference sensor selected for stability)

2. Microcontroller

OCASS Board:

• Data collection every 15 seconds (Timer 6)
• Signal processing and conditioning
• Actuator control via downlinks
• Embedded C programming

Interfaces used:

• ADC for LM94021
• I2C for BME680
• LoRaWAN for wireless transmission

3. LoRaWAN Transmission

Technical Choice:

• Range: ~10 km in rural environment
• Low energy consumption
• Reasonable cost
• The Things Network (TTN) for reception

Alternatives studied:

• WiFi rejected: limited range (~10m), high consumption, interference-sensitive

4. Cloud Platform - Datacake

Two developed dashboards:

Simplified dashboard:

• Quick visualization of essential data
• Manual actuator control buttons
• Fast actions in emergency

Complete dashboard:

• Historical graphs for trend analysis
• All environmental metrics
• Automation rules configuration

5. Automation & Actuators

Automated actions (Datacake Rules):

• Heating: Activation if temperature < threshold
• Ventilation: Trigger if poor air quality
• Irrigation: Activation if humidity too low
• SMS Alerts: Notifications for critical conditions

Manual controls:

• Individual buttons per actuator
• Emergency button for global shutdown

Testing & Validation

2-Phase Test Protocol

Phase 1 - LM94021 only:

• General operation validation with USART available
• Comparison of sent vs received data
• Downlink testing with indicator LED
• ✅ Transmission and reception validated

Phase 2 - BME680 Integration:

• I2C communication (replaces USART)
• Data consistency validation on platform
• Multi-parameter display verification
• ✅ All metrics operational

Results Obtained

Temperature sensors:

• LM94021: Stable, reliable, selected as reference
• BME680: Variations up to 3°C between measurements, constant 1°C offset
• → Decision: Use LM94021 for temperature

Other BME680 parameters:

• ✅ Humidity: Accurate and consistent
• ✅ Atmospheric pressure: Reliable
• ✅ Air quality: Usable for ventilation control

Automation:

• ✅ Downlinks received and processed correctly
• ✅ Actions triggered according to configured rules
• ✅ LED test validated (actuator proof of concept)

Impact & Optimization

Resource Savings

Reactive and adaptive system:

• Heating activated only when necessary
• Irrigation triggered based on actual needs (no waste)
• Ventilation optimized according to air quality
• → Significant reduction in water and energy consumption

Optimal Growing Conditions

• Continuous 24/7 monitoring
• Reactivity to environmental variations
• Historical data for analysis and predictions
• Alerts for critical conditions

Future Prospects

Technical Extensions

• Additional sensors: Soil moisture, sunlight, CO₂
• Artificial intelligence: Weather predictions and plant needs
• Field testing: Validation in real production conditions

Extended Applications

• Underground urban mushroom farms
• Urban green space management
• Specialized crops (aromatic herbs, microgreens)

Technologies Used

Hardware: BME680, LM94021, OCASS board, actuators (relays)

Communication: LoRaWAN (868 MHz), The Things Network (TTN), I2C, ADC, UART

Programming: Embedded C, LoRaWAN configuration

Cloud: Datacake (dashboards, alerts, API)

Skills Developed

• Embedded systems (C, timers, ADC, I2C, UART)
• IoT protocols (LoRaWAN, ABP, duty cycle management)
• Cloud architecture (TTN, Datacake, API)
• Environmental sensors and calibration
• Remote automation and control
• Energy optimization for autonomous systems