DEVELOPMENT OF AI-BASED MICROCONTROLLER SYSTEM DEPLOYABLE ON IOT PLATFORM ENHANCING RF PERFORMANCE AND ENERGY CONSUMPTION IN SMART ENERGY METERS

ABSTRACT The increasing demand for accurate and efficient energy monitoring has exposed the limitations of traditional power distribution systems, which rely heavily on manual meter readings and error-prone billing processes. Smart meter is an electric device with an energy meter chip for measuring the energy consumed and a wireless protocol for data communication and it has been introduced to address most of the issues. The accuracy in measurement and communication in smart meters through diverse means such as Programmed Meter Reading (PMR), which utilizes one-way communication, and Advanced Metering Infrastructure (AMI), which utilizes two-way communication to both transmit usage data and perform perception and maintenance. This study proposes the development of a smart energy meter powered by an AI-based microcontroller deployable on IoT platforms, which entails an AI-enhanced IoT-based smart energy meter that integrates machine learning (ML) for load forecasting, Non-Intrusive Load Monitoring (NILM) for appliance-level disaggregation, and optimized LoRaWAN-based RF communication for long-range, low- power data transmission. The system leverages an AI-driven microcontroller to analyze consumption patterns, predict demand, and detect anomalies in real-time. Additionally, energy-efficient IoT modules minimize the meter’s power consumption, ensuring sustainability. The system utilizes machine learning algorithms to classify and predict energy consumption patterns while optimizing RF communication for seamless data transmission. By integrating low-power IoT modules, the design ensures energy-efficient operation, reducing the overall power consumption of the meter itself. The proposed solution eliminates human intervention in meter readings, minimizes billing inaccuracies, enhances real-time energy monitoring for consumers and utility providers, improves RF signal strength up to 15 km in rural areas, and cost-effective upgrades for next- generation smart metering systems. Experimental results demonstrate improved RF signal reliability, integration of AI, and reduced energy waste through consumer energy awareness which also empowers utility providers with actionable insights, paving the way for a smarter, more sustainable energy ecosystem, making it a viable upgrade for modern and older power infrastructures.