Hammer A (2025)
Publication Type: Thesis
Publication year: 2025
Quantification of signals has been a practice in computer science for a long time. With sensors and sensor networks, all kinds of real world information can be mapped to data: Temperatures, colours, mass, speed and others. The desire to quantify such attributes of one's own body is a trend in itself, commonly called Quantified Self. Here, relevant signals often originate inside or around a person's body, for example heart rate, steps or the physical position. This data is voluntarily recorded by users and often stored for extensive periods of time, comparable to a digital, quantified diary. Such data is usually collected by personally owned, wearable devices like fitness trackers.
While this data is intended for personal analytics, sports training and, possibly, marketing purposes of the manufacturer, another purpose gains importance: Forensic investigations. The goal of forensic investigations is to find, and prove, the truth. This is an intrinsically objective goal, but in reality this is usually never fully reached. What matters is that facts and witnesses should be as objective as possible while simultaneously providing as much information as possible. Here, wearables step in as silent witnesses that diligently collect and store data of their owner with only minimal room for interpretation. As these devices are usually worn by people interested in precise, correct and dependable metrics, they act as valuable data providers for legal investigations. Nevertheless, these devices can also suffer from errors and manipulations, so it is vital to know the limits and check for signs that may restrict trustworthiness.
This thesis explores the topic of wearable witnesses, including device hardware and software, data extraction and analysis, interpretation of users' biological and geolocation data, and the importance of all these aspects in legal investigations. We have identified three aspects where this field of work is still lacking and have contributed accordingly. On the data extraction side, we formally define quality metrics for cloud storage forensic accessability and evaluate them for a number of common services. On the device side, we provide a definition and taxonomy for the term fitness tracker, compare it to related mobile devices, and introduce a common model for forensic analysis of fitness trackers. Finally, on the application data side, we discuss informational value and validity of geotraces and present a proof of concept implementation of a validation tool.
APA:
Hammer, A. (2025). Wearable Witnesses - Forensic Usage of Quantified Self Data (Dissertation).
MLA:
Hammer, Andreas. Wearable Witnesses - Forensic Usage of Quantified Self Data. Dissertation, 2025.
BibTeX: Download