Body Sensor Network: A Self-Adaptive System Exemplar in the Healthcare Domain

TL;DR

This paper presents SA-BSN, a self-adaptive body sensor network prototype designed for dynamic health monitoring, emphasizing reliability, energy efficiency, and flexibility in unpredictable healthcare environments.

Contribution

It introduces a novel self-adaptive system for healthcare sensor networks with configurable scenarios, noise injection, and an extensible controller, implemented in ROS.

Findings

Demonstrates effective self-adaptation in sensor networks

Balances system reliability and battery consumption

Provides a reusable framework for healthcare monitoring

Abstract

Recent worldwide events shed light on the need of human-centered systems engineering in the healthcare domain. These systems must be prepared to evolve quickly but safely, according to unpredicted environments and ever-changing pathogens that spread ruthlessly. Such scenarios suffocate hospitals' infrastructure and disable healthcare systems that are not prepared to deal with unpredicted environments without costly re-engineering. In the face of these challenges, we offer the SA-BSN -- Self-Adaptive Body Sensor Network -- prototype to explore the rather dynamic patient's health status monitoring. The exemplar is focused on self-adaptation and comes with scenarios that hinder an interplay between system reliability and battery consumption that is available after each execution. Also, we provide: (i) a noise injection mechanism, (ii) file-based patient profiles' configuration, (iii) six healthcare sensor simulations, and (iv) an extensible/reusable controller implementation for self-adaptation. The artifact is implemented in ROS (Robot Operating System), which embraces principles such as ease of use and relies on an active open source community support.