The fall detection application informs the consequences of falling, but it would better if we can prevent people, especially the elderly, from the accident. We introduce a fall prediction system which allows us to analyze the walking pattern of the user with cutting-edge sensors and a camera to model a normal pattern of the activity. Therefore, if the walking happens to be out of the normal pattern, the device informs us about the risk of falling. This system utilizes real-time signal and image processing procedures using AI-assisted algorithms based on machine learning and sophisticated deep learning neural networks, making our system more accurate. The device will be mounted on a belt worn by the user. This application assists you to care for your loved ones, noticeably lowering down the risk of their falling