Related Experiment Video
Updated: Aug 5, 2025
11:21
Methodology for Establishing a Community-Wide Life Laboratory for Capturing Unobtrusive and Continuous Remote Activity and Health Data
Published on: July 27, 2018
8.3K
An Innovative Random-Forest-Based Model to Assess the Health Impacts of Regular Commuting Using Non-Invasive Wearable
Mhd Saeed Sharif1, Madhav Raj Theeng Tamang1, Cynthia H Y Fu2
1Intelligent Technologies Research Group, ACE, UEL, University Way, London E16 2RD, UK.
Sensors (Basel, Switzerland)
|March 30, 2023
Summary
Commuting increases stress, raising blood pressure and altering brain waves (EEG). This study used biosensors to measure these effects, finding significant changes post-commute. Machine learning models accurately predicted stress levels.
Area of Science:
- Occupational Health
- Neuroscience
- Cardiovascular Health
Background:
- Regular commutes are linked to chronic stress, impacting physical and emotional well-being.
- Early recognition of mental stress is crucial for effective clinical intervention.
- Commuting stress can manifest through physiological and psychological changes.
Purpose of the Study:
- To investigate the health impacts of daily commuting using objective and subjective measures.
- To quantify the physiological and neurological effects of commuting stress.
- To develop predictive models for commuting-related stress.
Main Methods:
- Recruited 45 healthy adults across various commute types (bus, driving, cycling, train, tube).
- Utilized wearable biosensors for continuous electroencephalography (EEG) and blood pressure (BP) monitoring during commutes.
- Employed the Positive and Negative Affect Schedule (PANAS) for qualitative stress assessment.
- Applied correlation analysis and machine learning algorithms (Random Forest, SVM, Naive Bayes, KNN) for data analysis.
Main Results:
- Commuting significantly increased blood pressure and EEG beta wave activity.
- Positive affect ratings (PANAS) decreased post-commute (34.73 to 28.60).
- Systolic blood pressure was higher after commuting compared to before.
- EEG beta wave power exceeded alpha wave power after commuting.
- Random Forest model achieved 91% accuracy in predicting stress.
Conclusions:
- Daily commuting demonstrably impacts human health, inducing physiological stress responses.
- Wearable biosensor technology provides effective quantitative measures of commute-induced stress.
- Machine learning, particularly Random Forest, shows high accuracy in modeling and predicting commuting stress.
