🔌 Smart Plug IoT Project

This repository contains the implementation details for the Smart Plug, a third-year IoT project developed at Metropolia University of Applied Sciences by Mong Phan, Nadim Tran, Phuong Duong, and Xuan Dang, submitted on March 12, 2025.
A complete end-to-end IoT smart plug system, designed from scratch, from custom PCB layout to firmware development and final functionality testing. Built on Raspberry Pi Pico W and FreeRTOS, the system uses a modular, multi-tasking architecture with rich inter-task communication via queues, semaphores, and event groups to manage concurrent operations such as power sensing, MQTT communication, OTA updates, LCD display control, and safety enforcement. Features include real-time power tracking with ACS712, configurable power limits, a full-color local display, remote scheduling, and secure cloud interaction via MQTT. This project showcases a full-stack embedded design approach integrating hardware, firmware, cloud services, and front-end control.

📖 Project Report

The full project report is available at Smart_Plug_Project_Report_G7.pdf.

📝 Project Overview

The Smart Plug addresses limitations in existing commercial smart plugs, such as the absence of onboard displays, fixed power thresholds, and rigid interfaces. Built on the Raspberry Pi Pico W microcontroller, it integrates FreeRTOS for real-time task management, MQTT for secure cloud communication with the server, and an ACS712 current sensor for precise power monitoring. Key features include:

• 🔄 Remote Control: Toggle appliances on/off remotely via a web interface.
• 📊 Real-Time Energy Monitoring: Track voltage, current, and power consumption on a 1. TFT LCD display.
• ⚡ User-Defined Power Limits: Set adjustable power setpoints with hardware and software over-current protection (OCP at 4.5A).
• ⏰ Automated Scheduling: Configure on/off timers for efficient appliance operation.
• 💾 Persistent Storage: Store settings and energy data in I2C EEPROM across power cycles.
• 🌐 OTA Updates: Support firmware updates over-the-air (OTA) for maintenance.

The system uses a custom PCB, a cloud-based MQTT broker (HiveMQ), and a React.js front-end with a Python FastAPI back-end. It serves both practical consumer needs and educational purposes, enabling students to apply IoT concepts in an end-to-end solution.

🛠 Prerequisites

🔧 Hardware

• Raspberry Pi Pico W
• 1.54" SPI TFT LCD (ST7789 controller, 240x240 pixels)
• ACS712-05B current sensor (5A, 185 mV/A)
• STMicroelectronics M24256-BRMN6TP I2C EEPROM (256 Kbits)
• Microchip MCP79410-I/MS I2C RTC with battery backup
• 12V relay for switching
• Hardware OCP circuit (comparator, D-type flip-flop)
• 12VDC, 5A power adapter

💾 Software

• Firmware: C with FreeRTOS and Pico SDK
• Frontend: Node.js (v16+), React.js, Tailwind CSS
• Backend: Python 3.8+, FastAPI, Paho-MQTT

🚀 Usage

1. 🔌 Power Up:

   • Connect the smart plug to the 12V power supply. The LCD displays initial status (socket state, Wi-Fi, etc.).

2. 🌐 Connect to Wi-Fi:

   • Press the left button for 3 seconds to enter Access Point (AP) mode.
   • Scan the LCD QR code or connect to the plug’s Wi-Fi (IP: 192.168.4.1).
   • Enter router SSID and password via the web interface or QR code.

3. ⚙️ Configure Settings:

   • Setpoint: Use the web app to set a power limit (max 4.5A).
   • Schedules: Add timers (date, time, on/off) via the web app’s scheduling interface.
   • Toggle Socket: Press the right button or use the web app to turn the appliance on/off.

4. 📊 Monitor Energy:

   • View real-time voltage, current, and power on the LCD.
   • Check historical consumption (day/month/year) via the web app’s charts.

5. 🔄 OTA Updates:

   • Trigger firmware updates via the web app, which the OTA Task processes using MQTT and HTTP.

6. 🔔 Alerts:

   • The buzzer beeps for state changes, OCP triggers, or button presses.
   • The LCD shows “Socket: OL” for over-current events.

🧪 Testing and Validation

The system was rigorously tested for:

• 📏 Algorithm Accuracy: Load current (±2%), power consumption (±3%), and RTC drift (<1s/24h).
• 🎮 System Control: Socket toggle (100% success, <50ms), OCP response (<200ms), and timer accuracy.
• ⚡ System Performance: Stable over 48 hours, 120ms latency, 98% MQTT delivery rate, and ±1% energy reporting.

⚠️ Limitations and Future Work

• 🚫 Limitations:
  • Limited to 5A loads due to ACS712 sensor.
  • MQTT latency in unstable networks.
  • No integration with Alexa/Google Home.
• 🔮 Future Improvements:
  • Support higher current loads with upgraded sensors.
  • Add smart home ecosystem compatibility.
  • Enhance the web app with ML-based usage predictions.
  • Integration of multiple devices with different IDs

---

📌 Contributors

• Mong Phan
• Nadim Ahmed
• Xuan Dang
• Phuong Ta

---

⚖️ License

This project was developed for academic purposes at Metropolia University of Applied Sciences. User can freely use this project for educational purpose ONLY.