Ovladání chytré domácnosti pomocí radarových technologií

Abstract

This bachelor thesis focuses on the design and implementation of an autonomous smart home control system utilizing mmWave radar technology. The primary objective was to develop a reliable, privacy-respecting alternative to camera systems that is capable of detecting human micro-movements in real time. The sensory layer is based on a single-chip Texas Instruments IWR6843AOP radar operating in the 60 GHz band, which processes raw data directly on the chip using edge computing technology. A Raspberry Pi Zero 2W serves as the central control and communication unit. The core of this system is built on a fully asynchronous Python software architecture utilizing green threads (Eventlet). The system evaluates spatial data through a virtual zone state machine and dispatches non-blocking control commands to a KNX bus. Furthermore, the system integrates with Google Home via the MQTT protocol and the Smart NORA platform, incorporating a hybrid manual override logic to seamlessly resolve conflicts between automated sensing and explicit user commands. Practical experiments involving test subjects precisely confirmed the robustness of the solution, with the system achieving 100\% sensitivity in spatial detection and 100\% specificity in filtering noise and false alarms. Finally, a responsive web interface was developed using HTML5, CSS, and JavaScript for target visualization and system configuration.