PhysioLounge - eHealth for Optimization of Workplace Conditions

Working while sitting down for the majority of the day is becoming increasingly more prevalent in modern society, leaving workers at a higher risk for health issues related to sedentary behavior and the adoption of incorrect sitting postures. To reduce the absenteeism caused by these issues and optimize work productivity, necessity has risen to provide workers with a smart chair system for monitoring sitting posture, physiological signals and environmental conditions. Volunteers are needed to sit in a chair equipped with sensors for 15-30 minutes.

Motivation

For the past decades, economic advances and industrial innovation has lead to a gradual shift from manufacturing to service industries. This shift, combined with rapid technological developments, has resulted in an increasing number of people employed in sedentary occupations where workers spend the majority of their working day sitting. In fact, research shows office workers spend more than 70% of work hours sitting, placing them at a higher risk for health problems related to sedentary behavior (e.g, cardiovascular disease, obesity, etc.) and the adoption of incorrect sitting postures for prolonged periods of time (e.g. musculoskeletal disorders in the back, neck, and limbs).


Due to the absenteeism and loss of work productivity caused by these health issues, companies are faced with a substantial financial burden. So, how can employers optimize workplace conditions to promote the well-being of workers? Currently, the only tools available are the promotion of workplace health programs and sporadic health screenings. But, what if workers were provided with an instrument that would give them the ability to actively monitor their activity, health and workplace conditions on a day-to-day basis? This prospect is the focus of the Master’s thesis of ScientISST Leonor Pereira, which aims to develop an intelligent chair system – PhysioLounge -, capable of monitoring sitting posture, cardiovascular health, and environmental conditions found to have an impact on workers’ well-being and productivity.


Prototyping

To monitor sitting posture and other conditions, an ordinary office chair was modified to include a multitude of relevant sensors:


  • 3 load cells under the chair seat

  • 1 ECG sensor connected to conductive fabric on the arm rests

  • 1 temperature, humidity and air quality sensor

  • 1 light sensor


In this setup, a ScientISST SENSE was used to continuously acquire data from these sensors. The goal is to process this data into relevant information on sitting posture, cardiovascular health and workplace environment to be relayed to the user while they are working, so that they can be alerted to conditions conducive to poor health outcomes and productivity and take steps to improve on them.


Data on the weight applied to the 3 points of measurement in the chair seat is used to calculate the center of mass of the user on this plane. This feature is then used to classify their sitting posture using a machine learning model trained with similar data collected from a group of 10 participants, where they were instructed to adopt seven distinct sitting positions for one minute, repeating this process five times over the course of a week.


To assess cardiovascular health, the ECG signal is acquired from the conductive fabric on the arm rests when the user grabs them. However, this non-intrusive ECG acquisition is more prone to noise artifacts and lower signal-to-noise ratio, thus denoising algorithms and outlier detection methods were implemented to provide only high quality data.


All the data collected by the ScientISST SENSE is available in a Web App, accessible through any web browser on a PC or smartphone, which will also alert the user to: prolonged periods of time spent sitting; incorrect sitting postures; concerning ECG alterations; and environmental condition values (temperature, humidity, air quality and luminosity) outside of the recommended ranges by Occupational Health and Safety entities.


Participate in This Study

Now that the posture classification models have been trained, new healthy volunteers are required to test them. This experiment takes only 15-30 minutes, and the participants only have to sit on the PhysioLounge in the positions indicated by the examiner for one minute each. Easy, right?


If you are have some free time during the months of July and August and are eager to contribute to scientific advances, you can anonymously volunteer to participate by emailing leonor.pereira@tecnico.ulisboa.pt. Come and see how your usual sitting posture measures against the optimal position!

Project Gallery

Team of ScientISSTs

Technologies Used

true
Toolbox
true
CORE