Sebti, Abderrahim2026-10-282026-10-282026-06-27https://dspace.univ-tlemcen.dz/handle/112/27193The thesis proposes an intelligent UV-C sterilization system to reduce energy consumption and inefficiencies in maintenance of indoor air treatment. The device relies on a Raspberry Pi microcomputer and a pre-trained YOLO algorithm for real-time human occupancy detection and switch from continuous to dynamic and context-aware operation. A zero tolerance opto electronic interlock guarantees total optical containment of UV-C radiation. The system has two software interfaces: one for easy home use and one for tracking compliance in a multi zone healthcare setting. The empirical validation of the prototype resulted in an 80-90% reduction in bacterial colonies and a 66.6% reduction in active energy consumption. The project combines physical sterilization with intelligent algorithms to design a responsive and energy-efficient biomedical device and future iterations will mitigate initial aerodynamic limitations.enAir sterilizationUV-C radiationenergy optimizationartificial intelligenceEmbedded Systems.Thesis