University of Twente Student Theses

Login

Human-AI teaming for Conformity Assessment of Welded Joints: A Human Factors Perspective

Hof, M.J. (2022) Human-AI teaming for Conformity Assessment of Welded Joints: A Human Factors Perspective.

[img] PDF
1MB
Abstract:AI systems to assist weld inspectors are on the rise. In this research, a theoretical framework for human-AI teams was created to determine factors playing a role in human-AI interaction. Moreover, a study was conducted to test the effect of two independent variables (task division and AI accuracy) and one predictor variable (propensity to trust) on four outcome variables: Accuracy, efficiency, fear-induced stress, and the feeling of having a meaningful job (FoMJ). 18 weld inspectors of the company DEKRA were asked to interpret x-ray images of welds with the AI. An extension to the study was conducted with 6 different weld inspectors to determine a consensus for the correct interpretations. The parallel task division led to more accurate results. The sequential task division was more efficient, although large individual differences were present. There was no effect of the task division on fear-induced stress or FoMJ. Propensity to trust positively influenced FoMJ with history-based trust as a mediator. Since high accuracy is extremely important in weld inspection, the parallel task division should be favoured. The large individual differences in efficiency and the effect of propensity to trust on FoMJ illustrate that individual characteristics are important for successful implementation of an AI.
Item Type:Essay (Master)
Clients:
DEKRA, Utrecht, Nederland
Intergo, Amersfoort, Nederland
Faculty:BMS: Behavioural, Management and Social Sciences
Subject:77 psychology
Programme:Psychology MSc (66604)
Link to this item:https://purl.utwente.nl/essays/89356
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page