GRENZE International Journal of Engineering and Technology
Vol. 12
(2026), Issue 2
Vital Eyes – An AI-Powered Multimodal Fall Detection System
Authors
A Kavitha, Netra Srinivasan, Prathana Y, Raja Nandhini S
Abstract
India's elderly population has reached 140 million, and supporting independent aging is significant. Falls have become very common among the senior citizens, leading to undesirable situations and accidents. Existing devices provide limited support and lack immediate response, leading to imprecise decision making. Our system aims to resolve this by providing AI-powered emergency monitoring to mitigate after-effects of elderly falls. This system continuously monitors patient vitals, posture, sound classification and gesture detection and effectively makes a decision cumulating all the parameters, providing patient’s condition and immediately alerts when fall occurs. Through the MediaPipe algorithm, the machine learns abnormal gestures. It has three main significances: biomedical data analysis, audio detection of normal or abnormal sounds, and precise body monitoring using MediaPipe, which continuously provides the exact position. Based on these three factors, we can obtain a clear decision and determine the exact state of the patient. Whenever there is an abnormality, the system triggers an alert to the caretaker or the pre-registered responsive person. This enhances safety and strengthens the digital healthcare system. As the population rises, the care should also increase. As a future scope, it can incorporate real sensing and optimization for low-power hardware and enable large-scale deployment across homes and institutions. The ultimate aim is to create smart safety system and efficient fall detection to empower elderly citizens to live independently with confidence, dignity and security.
Pages:
268 - 275