Enhancing Academic Environments and IoT Education: A Dual-Purpose Smart Noise Management System Using Edge Computing

Modern universities require effective designs for Intelligent Learning Environments (ILEs) to enhance academic performance. This paper presents a dual-purpose ILE deployed in a university library that simultaneously improves acoustic conditions and provides hands-on training in Industrial Internet of Things (IIoT) technologies. The system employs a “Shift Left” edge computing paradigm with a Convolutional Neural Network (CNN) for real-time acoustic classification, running on an Advantech UNO-2271G V3 gateway. The study followed a 12-week evaluation involving a 4-week pre-deployment baseline and an 8-week post-deployment phase with 12 students from the School of Internet of Things. The environmental objective was successfully met, achieving a statistically significant 58.3% reduction in sustained disruptive noise events. Operational efficiency was confirmed with an average end-to-end latency of 178 milliseconds, satisfying the requirement for near-instantaneous intervention. Pedagogically, 91.7% of the students achieved mastery of complex IIoT tasks, including 8-bit model quantization and Modbus-to-MQTT protocol translation. These results demonstrate the ILE’s value as a live testbed for integrating cutting-edge edge AI challenges into advanced engineering education. Overall, the proposed system offers a verified model for enhancing learning spaces while simultaneously developing students’ practical technical skills.