Does my watch know that I need a break?

Almost everyone has one and we carry it with us all the time. Sometimes it knows more about us than we do ourselves. It helps us get from A to B and allows us to always have unlimited access to knowledge. It's all about the smartphone, of course. According to Statista, 95.5% of all 20- to 29-year-olds in Germany owned a smartphone in 2021. But the share of wearable users (e.g. smartwatches, fitness/health trackers) is also not to be scoffed at, at 30% in 2022, and the trend is on the rise. Wearing wearables not only makes it possible to document sporting activities or read news directly, but also to view physiological parameters (such as heart rate) continuously over a long period of time. By integrating different wearables and other information (such as data from digital questionnaires), many both exciting and challenging questions can be considered and current research can be integrated into everyday life. However, if you now want to record, combine and compare the data from different devices, the manufacturers of the devices often do not make it easy. The data is often pre-processed and made available in a manufacturer-specific app, as the following graphic illustrates:

That's why we developed the Android app SensorHub at the Chair of Digital Health - Connected Healthcare. With SensorHub, different wearables can be connected via Bluetooth, data can be recorded, labeled and uploaded to the cloud. 
 

SensorHub allows us to access raw data from sensors, run projects and develop machine learning models with extracted data. The app is modular, so it can be extended when new devices are needed for new projects. A backend running on a server provides long-term, centralized storage of configuration and sensor data.

In the current bachelor project, new functionalities for the SensorHub system are implemented and validated to support the so-called biofeedback. Roughly speaking, biofeedback is the reaction (e.g. in the form of haptics, acoustics or visuality) to previously defined vital parameters. For example, one can imagine that the smartwatch can remind one via vibration that it is time for a break during hours of programming, as it has been determined on the basis of physiological parameters (e.g. given by EEG measurements) that the concentration has rapidly decreased. Or, certain physiological measurements and derived metrics could be continuously visualized to enable better self-assessment and regulation of psychological and physical responses.