IoT-enabled Student Attendance System using RISC-V based Embedded System

Inefficiencies, cumbersome processes, and a significant risk of errors, delays, and misuse often plague conventional methods for tracking student attendance. To address these challenges, this paper introduces a smart monitoring system driven by the Artificial Intelligence of Things (AIoT), specifically designed for resource-limited embedded processors. This research examines a real-time facial recognition-based student attendance tracking system implemented on an edge device. Additionally, it assesses the effectiveness of ARM and RISC-V microprocessors as the foundation for a reliable edge computing platform. The system employs Python’s Dlib library and the ResNet model for effective facial recognition. Comparisons between the Raspberry Pi Model 4 B and the RISC-V StarFive VisionFive2 are made to determine the optimal platform for extensive deployments. The proposed edge system emphasizes security, automation, and reliability, facilitating real-time, precise attendance tracking and significantly improving operational efficiency for educators and administrators. The solution using the RISC-V processor is at least 10 times more energy-efficient than the competing solution based on the RISC processor instruction set architecture.