Research line: Learning and Innovation in Resilient Systems
PhD research project
Researcher: Bea Waelbers
Start 2021

The modern team-focused and service-oriented workplace abounds with frequent interpersonal interactions that can constitute major stressors. Particularly service employees who are constantly exposed to negative emotion expressions from customers have to pay the toll. 

An interdisciplinary team of the OU developed an artificially intelligent, automated stress assessment tool based on deep learning to overcome the limitations of employee
stress self-assessment. The deep neural network was trained on interpersonal stressors of service employees based on nearly 5,000 manually expert-coded customer service interactions and reached a balanced accuracy of 80% in real-time employee stress detection.

This project takes the next step in validating the tool: a validation against physiological measure of stress and an integration of intervention strategies based on the data. This is done in collaboration with the DHL call centre in Maastricht. An additional partner may be contacted. The physiological data will be collected via Fitbits and complemented with subjective questionnaire data. Based on theory, managerial insights and the data collected, in 
intervention strategies will be developed, integrated into the stress detection algorithm and assessed as to its effectiveness in the field. 

The project team

• Faculty of Management: dr. Alexander Henkel 
• Faculty of Sciences: dr. Stefano Bromuri, prof. dr. Marko van Eekelen