Towards Richer Insights in Human-Centred Observation Studies: Combining Thematic Analysis and Computer Vision
PUBLICATION DATE: 11 November, 2025 PUBLICATION AUTHOR/S: Yuan Liu, Glenda Amayo Caldwell, Markus Rittenbruch, Matthias Guertler, Muge Teixeira, Alan G BurdenDue to the precision and efficiency that collaborative robots (cobots) offer, they are becoming increasingly vital to advanced manufacturing, particularly for dynamic and complex tasks that depend on human intelligence, such as decision making. To support cobot adoption, this paper presents a novel method for analyzing human decision-making and task complexity using video observation. Moving beyond conventional video analysis, the proposed approach combines thematic analysis with computer vision-based motion recognition to reveal behavioral patterns and decision-making processes. Through the application in a real-world manufacturing gasket room task, we demonstrate how the integration of interpretive coding and computational motion data can uncover insights into task structure, decision points, and potential cobot intervention zones. This method contributes a practical tool for aligning cobot functions with human needs in complex settings. It demonstrated how generative intelligence can augment human-centered research, and inform the design of future collaborative system that aligned with planetary sustainability.
RELATED PROGRAM/S:Human-Robot-Interaction Publication link
View all publications