Integrating the interest of dancers in monitoring dance quality with the interest of engineers in monitoring loads and vibrations, this research studies floor vibrations induced by human activity while classifying and characterizing dances simultaneously. This research uses smart sensing capabilities with readily available low-cost Arduino sensors equipped with accelerometers. The paper describes the procedures to extract step features from the signal to categorically classify different dance steps. The major contribution of this work is to demonstrate that structural vibration can be used to classify dance steps and provide meaningful information about the harmony of the dances. The conclusion of this research accomplishes that using this new Cyber-Physical Systems (CPS), dancers’ performance can be objectively classified using floor vibration data.