Exploring the core cognitive dynamics of MR-enabled in-situ assembly visualization

Blind area manual assembly arises from the narrow and inter-obscuring structures of products which makes it difficult for workers to see the assembly site. Mixed Reality (MR) visualization systems can provide intuitive interfaces to improve efficiency and reduce errors in blind assembly, but the neural mechanisms in MR visualization of blind area assembly are not well understood. We conducted a user study (N = 24) that compared the difference in cognitive load between using a future advanced visualization system simulated using MR (Low latency, high accuracy and easy detection of current movements) and blind area assembly with the intention of explaining its core dynamics. Significant differences in electroencephalographic (EEG) features including power spectral density (PSD), coherence (Coh) and event-related potentials (ERPs) were analyzed between the two conditions. And the overall EEG was divided into three segments based on three events (Pickup, Localization and Insertion). The results showed that MR visualization exhibits PSD features such as decreased theta power, increased gamma power, and increased beta power, Coh features such as weakened theta, alpha connectivity, and increased beta, gamma connectivity, and shorter latencies for ERP components such as N2 and P300 at specific phases. We discuss these insights and directions for future work. We hope that research this may provide suggestions for the construction of current blind area assembly visualization systems.