A Hybrid AR-IoT Architecture with AI-Enabled Interaction Measurement for Smart Switchboard Control

Abstract

Smart device control is becoming increasingly embedded in homes, laboratories and intelligent environments. However, conventional physical switches, mobile dashboards and voice-based interfaces may limit interaction visibility, flexibility and feedback reliability. Augmented Reality (AR) offers a more user-friendly control layer, but most of the AR-IoT implementations available today are restricted to virtual switching demonstrations, rather than having quantifiable interaction, physical actuation, and synchronized device-state verification. The proposed study offers a quantifiable hybrid AR-IoT architecture for smart switchboard control that integrates marker-based AR interaction, IoT communication, ESP8266/ESP01-based device actuation, feedback synchronization, and AI-enabled interaction measurement within a unified control loop. A working prototype was developed using Unity, Vuforia, Blynk and embedded IoT hardware to facilitate representative control of lights, fans and appliances. Simulation-based validation was done for latency, command reliability, marker detection, feedback synchronization, multi-modal intent recognition, intelligent decision performance, anomaly detection, and scalability. Results showed that the architecture achieved 357 ms average command-to-feedback latency, 96.00% command success rate, 97.33% feedback synchronization accuracy, 94.33% multi-modal intent recognition accuracy, and 94.55% AI-enabled system performance index. These results demonstrate that the proposed architecture pushes the boundaries of AR-IoT control beyond virtual switching by integrating measurable physical actuation, interaction intelligence, and device-state verification, to provide a practical basis for small-scale smart home, laboratory, and instrumentation applications.