Edge AI-Based Intelligent Controller for Energy and Water Efficient Smart Washing Machines

Abstract: Conventional washing machines rely on rigid, pre-programmed cycles that ignore the actual physical state of the laundry load, leading to substantial and unnecessary water and energy waste [1]. To tackle this everyday inefficiency, this paper presents a practical, smart washing machine controller powered by Edge AI and Internet of Things (IoT) technologies [2], [3]. The proposed system utilizes an ESP32 microcontroller outfitted with a comprehensive sensor suite, continuously monitoring load weight, multi-axis vibration, real-time water flow, temperature, and power consumption [4]. Rather than relying solely on high-latency cloud processing, the device performs local data acquisition and preprocessing at the edge, securely transmitting refined metrics to a Flask server via a REST API [5]. A Random Forest machine learning model utilizing an ensemble of 200 decision trees dynamically analyzes these variables to predict the optimal water volume, wash duration, and spin speed for each unique load. Live system metrics and operational states are continuously visualized using the ThingSpeak IoT platform. Experimental validation confirms that this adaptive, real- time approach significantly outperforms conventional fixed-cycle methods. The proposed Edge AI controller achieved a 24% reduction in water usage, an 18% drop in overall energy consumption, and shortened wash durations by 25%. Furthermore, the Random Forest model demonstrated robust predictive accuracy, achieving an 𝑅2score of 0.87. Ultimately, this framework offers a highly practical, retrofit-friendly solution for transforming standard household appliances into sustainable, genuinely intelligent systems.

Keywords: Edge AI, Internet of Things (IoT), Random Forest Regression, Smart Home Appliances, Resource Optimization, Real-Time Data Acquisition, Sensor Integration, Predictive Modeling, Cloud Data Visualization