Does One-Size Training Fit All? Evaluating Adaptive Learning for VR Assembly Training

Abstract

Virtual Reality (VR) is gaining popularity and is increasingly adopted across various industries for its potential to deliver immersive and effective skill development. However, we observe that VR training often follows a one-size-fits-all approach. Trainings typically do not adapt to to individual skill levels, which is particularly important in industrial assembly, where user profiles and expertise levels vary widely. To address this, we applied the concept of adaptive learning to VR assembly training, enabling the system to dynamically provide assistance levels when users struggle and gradually reduce support as their proficiency increases. This paper investigates the learning performance and subjective impact of two types of such adaptive approaches and a non-adaptive variant in a VR user study with 36 participants. The results show that adaptive training significantly enhances user experience and reduces perceived workload. At the same time, adaptive VR learning is found to have a positive impact on learning performance (quantified as a reduced number of assembly mistakes after training). In summary, our findings underscore the potential of applying adaptive learning approaches in VR. To guide future research, we propose guidelines to support the practical adoption of adaptive learning in VR training in manufacturing and beyond.