LeafLink: Smart Plant Care System

Project Overview

LeafLink is an IoT-enabled plant care system designed to monitor environmental conditions and automatically water indoor plants when needed. Combining embedded sensors, a servo-actuated valve, MQTT-based data publishing, and a custom 3D-printed structure, the system continuously tracks plant health while reducing the need for manual watering. The system monitors and publishes the soil moisture, atmospheric temperature, air humidity, and illuminance.

Key Results

Personal Contributions

• Led the mechanical design, embedded software development, and system integration of the project.
• Designed and iterated a compact 3D-printed structure integrating the reservoir, valve, sensors, electronics, and wiring.
• Developed the Arduino software responsible for environmental sensing, valve actuation, and MQTT-based IoT communication.
• Performed calibration, debugging, deployment testing, and multi-day system validation.

Mechanical Design

The mechanical design was developed as a compact clamp-on attachment rather than a replacement plant pot, allowing the system to integrate with existing indoor planters. The 3D-printed structure housed the reservoir, valve, servo motor, sensors, and electronics within a single assembly while minimizing water exposure to sensitive components.

Multiple design iterations refined component packaging, reservoir integration, and electronics placement. The final design provided reliable valve actuation, accessible electronics, and secure mounting while maintaining a compact footprint suitable for indoor use.

Exploded Assembly Showing Integrated Components

Design Iteration

Electronics and System Architecture

• ESP32-based microcontroller for sensor reading, servo control, and IoT communication
• Soil moisture sensor for determining when the plant required water
• DHT sensor for atmospheric temperature and air humidity monitoring
• Photocell sensor for measuring plant light exposure through illuminance
• Servo motor for opening and closing the mechanical water valve
• MQTT publishing system for sending sensor readings to the Stevens IoT network

The system architecture centered on an ESP32 microcontroller that monitored environmental conditions, evaluated soil moisture levels, controlled valve actuation, and published sensor data to the Stevens IoT network. Special consideration was given to separating water-handling components from the electronics while maintaining accessibility for testing and maintenance.

LeafLink Mounted to an Indoor Planter

Software and Control Logic

The control software monitored soil moisture, temperature, humidity, and illuminance while continuously evaluating whether watering was required. When soil moisture dropped below a predefined threshold, the servo motor opened the valve to deliver water before returning to its closed position.

Environmental data was published to the Stevens IoT network every 10 seconds using MQTT, enabling remote monitoring and multi-day deployment testing. System reliability was improved through iterative debugging, including resolving communication issues that initially prevented sensor data from publishing online.

Challenges and Engineering Decisions

Future Improvements