The growth of the elderly population intensifies the demand for effective and discreet fall detection systems that preserve user privacy. Radar-based and Internet of Things (IoT) technologies stand out in this context for offering high accuracy, continuous operation and security. Among the available solutions, the CBLSTM model, applied to Doppler radar spectrograms, has an accuracy of over 98%. Frequency Modulated Continuous Wave (FMCW) millimeter wave radars integrated with the IoT enable the generation of point clouds, local data processing and automatic alerts to caregivers, for example. This work explores how the combination of artificial intelligence, radar and IoT can enable scalable, reliable and non-invasive systems for elderly care.