The Noise of our Daily Motion: General Spectral Characteristics of Human Mobility and Activity

In recent decades, a strong focus has emerged on exploring the scale-independent nature of our daily spatial motion. Similarly, heavy-tailed distributions have been observed for human locomotor activity, measured by actigraphs in medical fields. We recently proved that the raw acceleration data and also the activity signals calculated from them in diverse ways exhibit a general spectral characteristic; 1/f noise is observable above a certain cutoff frequency, while white noise and peaks corresponding to daily rhythms are visible at lower frequencies. We show that this pattern is strikingly similar to what we found earlier for GPS data and this similarity raises fundamental questions like what is the relation between human mobility and physical activity, what are the benefits of the analysis from this perspective, and how it helps us to better understand and model the scale-independent nature of human dynamics.