AI Smart Hydroponics Monitoring and Nutrient Prediction System

This is an excellent industry-level final year engineering project because it combines IoT + AI + ESP32 + Automation + Cloud + Agentic AI, matching current Industry 4.0 and Smart Agriculture trends. AI Smart Hydroponics Monitoring and Nutrient Prediction System Complete Project Title AI Smart Hydroponics Monitoring and Nutrient Prediction System using ESP32, Agentic IoT, n8n Automation, Telegram Voice Alerts, Google Sheets, ThingSpeak Cloud, and AI-Based Nutrient Prediction Project Overview Hydroponics grows plants without soil by supplying nutrient-rich water directly to plant roots. Traditional hydroponic farms require constant monitoring of pH, EC, water level, temperature, humidity, and nutrient concentration. This project automates the entire monitoring process using an ESP32-based IoT system. Sensor data is uploaded to the cloud, analyzed by an AI agent, logged to Google Sheets, visualized on ThingSpeak, and used to generate Telegram alerts and voice notifications whenever abnormal conditions are detected. The AI module predicts nutrient requirements based on environmental conditions and historical data, helping optimize plant growth while reducing manual intervention. Objectives Monitor hydroponic water quality in real time. Predict nutrient requirements using AI. Automatically notify users of abnormal conditions. Upload sensor data to cloud platforms. Maintain historical records in Google Sheets. Display live dashboards using ThingSpeak. Automate workflows using n8n. Enable remote monitoring through Telegram. Hardware Components Component Quantity ESP32 DevKit V1 1 pH Sensor 1 EC Sensor 1 DS18B20 Water Temperature Sensor 1 DHT22 Temperature/Humidity Sensor 1 Water Level Sensor 1 TDS Sensor 1 Float Switch 1 Relay Module (4 Channel) 1 Water Pump 1 Nutrient Pump A 1 Nutrient Pump B 1 Air Pump 1 LCD/OLED Display 1 Buzzer 1 LED Indicators 3 Power Supply (12V/5V) 1 Software Requirements Arduino IDE ESP32 Board Package ThingSpeak Google Sheets n8n Telegram Bot PHP MySQL HTML CSS JavaScript Sensor Functions pH Sensor Measures acidity or alkalinity. Ideal range: 5.8–6.5 EC Sensor Measures nutrient concentration. Ideal: 1.2–2.5 mS/cm TDS Sensor Measures dissolved nutrients. Water Temperature Ideal: 18–24°C Air Temperature Ideal: 22–28°C Humidity Ideal: 50–70% Water Level Ensures sufficient nutrient solution. System Architecture Sensors ↓ ESP32 ↓ WiFi ↓ Cloud Server ↓ ThingSpeak Google Sheets PHP Database ↓ AI Agent ↓ Prediction ↓ n8n Workflow ↓ Telegram ↓ Voice Notification Working Principle Step 1 Sensors continuously measure pH EC TDS Temperature Humidity Water Level Step 2 ESP32 reads all sensors every few seconds. Step 3 ESP32 uploads data ThingSpeak PHP Server Google Sheets Step 4 AI Agent checks Normal High EC Low pH Low Water High Temperature Low Nutrient Step 5 If abnormal ↓ n8n Workflow ↓ Telegram Notification ↓ Voice Alert ↓ Store History Complete Circuit Connections pH Sensor VCC → 5V GND → GND OUT → GPIO34 EC Sensor OUT → GPIO35 Water Temperature DATA → GPIO4 DHT22 DATA → GPIO15 Water Level Signal → GPIO32 Relay Module Pump → GPIO26 Nutrient Pump A → GPIO27 Nutrient Pump B → GPIO14 Air Pump → GPIO25 OLED SDA → GPIO21 SCL → GPIO22 Flowchart START ↓ Initialize ESP32 ↓ Connect WiFi ↓ Initialize Sensors ↓ Read Sensors ↓ Display Values ↓ Upload Cloud ↓ AI Prediction ↓ Normal? ↓ YES ↓ Continue ↓ NO ↓ Relay Control ↓ Telegram Alert ↓ Voice Alert ↓ Google Sheets ↓ ThingSpeak ↓ Repeat ESP32 Program Modules WiFi Manager Sensor Module OLED Display ThingSpeak Upload HTTP Client Relay Controller AI Prediction Telegram Client Google Sheets Client OTA Update AI Nutrient Prediction Logic Inputs pH EC Temperature Humidity Water Level Previous Nutrient Data Example Rules IF pH < 5.5 ↓ Add Base Solution ------------------- IF EC < 1.2 ↓ Add Nutrient Solution ------------------- IF Temperature > 30°C ↓ Turn Cooling Pump ON ------------------- IF Water Level Low ↓ Turn Pump OFF ↓ Notify User ------------------- IF Humidity < 45% ↓ Turn Fogger ON AI Decision Table Condition AI Action Low pH Add Base High pH Add Acid Low EC Nutrient Pump A High EC Add Water Low Water Stop Pump High Temp Cooling Pump High Humidity Exhaust Fan ThingSpeak Fields Field1 pH Field2 EC Field3 Temperature Field4 Humidity Field5 Water Level Field6 TDS Field7 Pump Status Field8 Prediction Google Sheets Columns Date Time pH EC Temperature Humidity Water Level TDS Pump Prediction Remarks Telegram Notifications Example ⚠ Hydroponics Alert pH = 5.2 Low Nutrient Adding Solution A Time: 10:42 AM Voice Alert Attention! Hydroponics nutrient level is low. Automatic dosing has started. Please verify the tank. n8n Workflow Webhook ↓ Receive ESP32 Data ↓ IF Condition ↓ AI Analysis ↓ Telegram ↓ Google Sheets ↓ ThingSpeak ↓ Store Database ↓ Generate Voice ↓ Finish PHP Dashboard Dashboard contains Login Home Live Sensor Data Historical Graph Pump Control AI Prediction Reports Settings Export CSV User Management Database Tables Users id name email password SensorData id ph ec temperature humidity tds waterlevel prediction timestamp Alerts id message status time AI Agent Features The AI agent: Monitors every sensor reading in real time. Compares readings with optimal crop thresholds. Predicts nutrient deficiency trends. Suggests corrective actions. Triggers automation rules through n8n. Learns from historical data to improve recommendations. Deployment Guide Phase 1: Hardware Assemble the hydroponic setup. Connect all sensors and relay-controlled pumps to the ESP32. Power the system with a regulated 5 V/12 V supply. Phase 2: Firmware Install the ESP32 board package in Arduino IDE. Add required libraries (WiFi, HTTPClient, OneWire, DallasTemperature, DHT, ArduinoJson, etc.). Configure Wi-Fi credentials, API keys, and calibration constants. Upload the firmware. Phase 3: Cloud Create a ThingSpeak channel with the required fields. Create a Google Sheet and expose an Apps Script Web App endpoint. Set up a PHP/MySQL server to receive and store sensor data. Phase 4: Automation Create a Telegram bot with BotFather and obtain the bot token. Import the n8n workflow. Configure Telegram, Google Sheets, HTTP, and AI nodes with your credentials. Test end-to-end data flow. Phase 5: AI Start with rule-based predictions. Collect historical data. Train a regression or classification model (e.g., Random Forest or XGBoost) to predict nutrient dosing based on environmental conditions. Replace or augment the rule engine with the trained model. Phase 6: Testing Calibrate sensors. Verify cloud uploads. Simulate abnormal conditions (low pH, low water level, high temperature). Confirm pump activation, Telegram messages, voice alerts, and dashboard updates. Future Enhancements AI-based crop growth prediction. Automatic nutrient dosing using peristaltic pumps. Camera-based leaf disease detection with ESP32-CAM. Computer vision for growth-stage analysis. Weather forecast integration to optimize greenhouse conditions. Mobile app for Android and iOS. Multi-zone hydroponic farm management. LoRaWAN connectivity for large farms. Solar-powered operation with battery backup. Digital twin of the hydroponic farm for simulation and optimization. Suggested Project Deliverables For an engineering submission or startup prototype, you can package this project with: 200–250 page project report. IEEE-format research paper. Seminar presentation (PPT). Complete ESP32 firmware (Arduino IDE). PHP + MySQL web application. HTML/CSS/JavaScript frontend. MySQL database schema. Professional circuit schematic. PCB layout (KiCad or EasyEDA). Complete n8n workflow (JSON). Telegram bot integration. Google Sheets integration. ThingSpeak dashboard configuration. AI nutrient prediction module. Testing report with results and screenshots. Viva questions with answers. Installation, maintenance, and deployment manual. This scope is suitable for a major final-year engineering project, an IEEE publication, and as the basis for an AgriTech startup prototype.