Mobile forensics: A smartphone-based activity logger

Nowadays, mobile communication devices allow ubiquitous access to a plethora of rich contents and services, intended to make life easier. As a consequence, almost everyone carries a mobile phone almost everywhere. These devices can unobtrusively capture data from a multitude of built-in sensors, which may help detect and keep a record of relevant activity taking place. This is the starting point of this contribution - can a smartphone motion activity analysis and logger collect data useful for forensics applications? In fact, many daily problems could benefit from such an application, as the case of physical bullying, among many others. If the suspected activity is timely detected, then preventing alerts can be automatically generated, contributing to diminish the incidence of the problems. In any case, by keeping record of the detected events, and eventually by activating additional sensors (e.g., microphone, camera, GPS) to gather additional information, one might consider its latter usage for proving the event occurred at a given time and location. This paper proposes a smartphone application capable of identifying a set of user activities and inferring a semantic interpretation that could be useful for forensics analysis. It relies on smartphone sensors such as the three-axis accelerometers, microphone and GPS. The paper also identifies a list of application scenarios, reflecting some distress situations in which the device can keep record of relevant detected events for later proving they occurred according to given timeline.