IoT-Motion Sensor Device and Data Analysis : (Motion Detection and Step-Count Algorithm)

In the last one decade, with the advancement and developmental changes in artificial intelligence (AI) technology and the usage of smart-phones, motion-detection and step-count applications have increased the public attention. Because they are playing a significant role in the fields of device positioning, energy harvesting and facial and behavior recognition etc. These motion-detection sensors and smart-phone applications are playing a vital role in human's life. In this article, we are suggesting a novel peak detection motion algorithm to instantaneously detect foot movement and steps are counted by using Nordic Thingy device as a motion sensor by holding an android phone (Note 5A) in hand to save the walk data. The sensor device and unconstrained cell phones would have the ability to place not only arbitrary position but also alterable position. Depending on the foot-step motion periodicity and gyroscopes sensitivity; the projected algorithm extracts the features of time domain from a three dimensional (3D) angular velocity of a cell-phone with the help of Fast Fourier Transform (FFT) and classifies whether the sensor device with the smartphone of user sensed walk or not regardless of its placement or location of Thingy-sensor. The proposed algorithm is developed and experienced in real time environments for walking, running, climbing up and down stairs. It indicates promising results for walking at almost 99% accuracy. Contrasting, the conventional motion detection and step-count approaches, the errors are fixed in proposed algorithm and long distances do not affect its performance.