Skip to main content
To KTH's start page To KTH's start page

Ergonomic Risk Assessment and Intervention through Smart Workwear Systems

Time: Fri 2019-12-06 13.00

Location: Lecture hall T1, Hälsovägen 11C, Huddinge (English)

Subject area: Medical Technology

Doctoral student: Liyun Yang , Ergonomi

Opponent: Professor Karen Søgaard, University of Southern Denmark

Supervisor: Professor Jörgen Eklund, Ergonomi; Professor Mikael Forsman, Ergonomi, Karolinska Institutet

Export to calendar


The rapid development of wearable technology has provided opportunities to ergonomics research and practice with new ways for workload measurements, data analytics, risk assessment and intervention. This thesis aims at developing and evaluating methods using wearable technologies to assess physical risk factors at work, and further to give feedback to employees to improve their work techniques.

One smartphone application (ErgoArmMeter) was developed for the assessment of upper arm postures and movements at work. The application uses integrated signals of the embedded accelerometer and gyroscope, and processes and presents the assessment results directly after a measurement. Laboratory validation with 10 participants was performed using an optical tracking system as standard measurement. The results showed that the application had similar accuracy compared to standard inclinometry for static postures and improved accuracy in dynamic conditions. With its convenience and low cost, the application may be used by researchers and practitioners in various scenarios for risk assessment.

Three models for assessment of work metabolism (WM) using heart rate (HR) and accelerometers (ACCs) were evaluated during simulated work tasks with 12 participants against indirect calorimetry as standard measurement. The HR + arm-leg ACC model showed best accuracy in most work tasks. The HR-Flex model showed a small bias for the average of all tasks. For estimating WM in the field using wearable technologies, the HR-Flex model or the HR + arm-leg ACC model may be chosen depending on the need for accuracy level and resource availabilities. Further improvement of the classification algorithm in the HR + arm-leg ACC model is needed in order to suit various types of work.

Two smart workwear systems were developed and evaluated. Smart workwear system 1.0 consisted of a sensorized vest, an inertial measurement unit (IMU) and an Android tablet application. It assessed risks of high physiological workload and prolonged occupational sitting/standing. The results were visualized by color-coded risk levels. The system was evaluated with 8 participants from four occupations in a field study. It was perceived as useful, comfortable and not disturbing by most participants. Further development is required for the system for automated risk assessment of various ergonomic risk factors in real work situations.

Smart workwear system 2.0 consisted of an instrumented t-shirt with IMUs, vibration units and an Android smartphone application. It provided vibrotactile feedback to users’ upper arm and trunk when predefined angular thresholds were exceeded. The system was evaluated for work postures intervention in industrial order picking among 15 participants. It showed to be effective in improving the trunk and dominant upper arm postures. The system was perceived as comfortable and useful. The vibrotactile feedback was evaluated as supportive for learning regarding workplace and task design among the participants.

In conclusion, the research in this thesis showed that wearable technologies can be used both in the laboratory and field for assessment of physical risk factors at work and intervention in work technique improvement. With further research and development, smart workwear systems may contribute to automated risk assessment, prevention of work-related ill health, and improvement of the design and overall quality of work.