Stress changes our behaviour – technology embedded in the work environment can identify individual signs of stress

Blog post
Johanna Kallio
Kasvokuva, jossa tummahiuksinen nainen,  tohtori Johanna Kallio, katsoo kameraan.

Having stress is part of life, but can unobtrusive measuring of long-term stress levels be used for increasing employee well-being?

Short-term stress boosts performance and helps us cope in challenging situations. However, stress may wear us out if we lose the feeling of control, the stress continues and we do not have enough time for recovery.  Being under prolonged stress also ages our cells and increases the risk of multiple diseases, says Nobel laureate, Dr Elizabeth Blackburn.

Long-term stress has become a major work-related problem in developed countries. According to an EU study, 24% of employees experience stress in their work on a continuous basis and 35% feel exhausted at work. Work-related stress also generates significant costs. According to the study, up to half of the workdays lost are caused by work stress. In order to avoid causing excessive stress to employees and to provide support for them, we should be able to recognize when someone is in a continuous state of stress and which factors are causing it.

According to the study, up to half of the workdays lost are caused by work stress.

Wearable smart devices interpret the body's signals

Surveys are the traditional method for measuring perceived coping at work, but filling out and analysing surveys takes time, so they are not suited for continuous monitoring of well-being at work. 

Commercial wearable technology applications, such as smart wristbands or rings, can measure the user's physiological signals in everyday life. For example, the heart rate variability indicates that the workday has been stressful and that the employee needs time for recovery. Armed with this information, the employee can then try to adjust other stress factors in his or her life, to keep the overall loadat a manageable level. 

However, the use of wearable devices requires commitment, and they cannot always be worn in work situations. Therefore, technology embedded in the work environment unnoticeable to the user has become an object of active research and development.

Stress changes behavioural patterns

Stress reflects on the way employees behave physically, socially and digitally. Perceived stress can be observed, for example, by using motion detectors embedded in a workspace or sensors measuring indoor environmental quality. The data thus produced is analysed for changes in the employees’ motions or postures and in the ambient conditions of the workspace.

Studies have also shown that stress changes the way people use digital devices, such as smartphones, tablets, or computers. In practice, the digital behaviour is measured using software installed on the device, which analyses how and for how long we use different types of applications, and our writing tempo or keystroke rhythms. An employee's state of stress may manifest itself, for example, as an atypical keystroke rhythm or duration of application usage.

A common denominator to the methods that apply technology embedded in the working environment is the use of machine learning. Whether the measured object is about physical movements, smart device usage or the environment, the data is modelled using algorithms and the state of stress experienced by the employee is identified as a deviation from normal behaviour.

Detecting stress is challenging because every individual is unique. For example, I manifest work stress by moving less than I normally would at my workstation, while my lively colleague's motions increase in a similar overload situation. In addition, the rush and time pressure are reflected in unique ways on the way people use their smartphones or computers. Therefore, most accurate identification is achieved not by using general stress models based on multiple people's data but by using individual modelling methods for stress.

A common denominator to the methods that apply technology embedded in the working environment is the use of machine learning.

Aiming at comprehensive well-being at work

The Mad@Work project, led by VTT, develops reliable methods for long-term monitoring of the stress level of knowledge workers, where the demands of privacy, information security and legislation are taken into account. The goal is to support employees' ability to cope at mentally straining work and to enable early-stage identification of the individual factors that cause stress. Timely support measures aimed at reducing employees' stress factors, increasing their personal resources and improving the working culture prevent burnout. In the coming years, investments in employee well-being will be visible as increased productivity at the level of individuals, organisations and society alike.

Share