An open-source sensor platform for analysis of group dynamics

TL;DR

This paper presents an open-source, modular firmware for wearable badges used in group behavior analysis, featuring efficient data storage, platform-independent data exchange, and power optimization, applicable to various projects.

Contribution

It introduces a new firmware architecture with modular design, combining data storage, serialization, and synchronization techniques, enhancing accuracy and power efficiency for wearable sensor platforms.

Findings

Significant reduction in power consumption.

Modular and extensible firmware architecture.

High accuracy in behavior quantification.

Abstract

The collaboration of several people in groups is becoming more and more important nowadays. Teamwork is often used for decision-making processes and for solving complex problems. Research in this area focuses on the quantification and analysis of behavior within and between groups. By using wearable electronic devices, such as badges, this quantification can be performed automatically and with high accuracy. The goal of this work is the design and implementation of a new firmware for the badges of the Rhythm project - an open-source project of the Human Dynamics Group of the MIT Media Lab. The firmware is characterized by a modular and extensible architecture and combines different techniques. These include a filesystem, which efficiently stores sequentially generated data based on a virtual memory abstraction, a serialization library, which enables a platform-independent exchange of structured data, and a time synchronization technique, which compensates frequency deviations of the oscillator. In addition, an automated test environment was designed for the verification of individual functional software components of the application. As a result of detailed analysis and special measures, the power consumption could be significantly reduced compared to the previous implementation. Due to the hierarchical and modular structure and the high degree of abstraction, the developed techniques can also be integrated into other projects and platforms.