An Innovative IoT-Based Automation System for Sustainable Textile Wastewater Recovery and Energy Optimization

In this study, we developed an innovative and eco-friendly automation system to help recover wastewater from the textile industry, aiming to reduce its environmental footprint. The backbone of our setup is a Raspberry Pi 5 microcontroller, which is paired with several sensors that monitor pH (with an accuracy of ±0.1), temperature (DS18B20, accurate to ±0.5°C from -55°C up to +125°C), dissolved oxygen (0–20 mg/L), conductivity (0–20,000 μS/cm) and voltage. For three weeks, we ran experiments using the electrocoagulation method. Under optimal conditions—specifically, at pH 7 and a voltage range of 15–20 V—we achieved removal rates of over 94% for colour and turbidity, and more than 90% for chemical oxygen demand (COD). We collected sensor readings every six seconds and stored this information in real-time to both a Firebase Realtime Database and a local CSV file via Wi-Fi. When we analysed the results using machine learning models, XGBoost delivered the best performance, with an R2 value of 91.3%. We used Python-based Firebase SDK components to ensure that all data was transmitted securely and reliably to the cloud. The system also features a user-friendly web dashboard, allowing users to monitor operations remotely and in real time. To meet the system's energy needs, we relied on 3.51 kWp TOPCon photovoltaic panels, which generated 12–14 kWh of electricity daily. Looking ahead, we plan to further optimize solar energy production using advanced machine learning and deep learning algorithms. Depending on the operating conditions, energy consumption ranged from 13.33 to 240 kWh per cubic meter, but we were able to keep it around 40–50 kWh/m3 under optimal settings. All in all, this integrated system offers a practical and sustainable solution for managing industrial wastewater, and it takes a step forward in bringing digital transformation and environmental responsibility to the field.