Smart wristband has recently emerged as one of the most popular smart wearable devices. With the help of smart wristbands, we can easily gather and record the user's data of everyday life including exercising, sleeping, diet, etc. In this work, we explore the feature of falls of the old and redesign the smart wristband to detect their movement. We propose a novel algorithm to judge whether he/she stops in a sudden, then produce vibrations through the man-machine interactive interface that provided by the smart wristband. The wearer can judge whether it is misjudged if he is still in conscious, if so he can press a button to stop the shaking or the wristband will give an alarm through communication interface to the medical care center.