Cognitive Workload of Tugboat Captains in Realistic Scenarios: Adaptive Spatial Filtering for Transfer Between Conditions

Miklody, Daniel; Blankertz, Benjamin · 2022 · Crossref

DOI: 10.3389/fnhum.2022.818770

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study addresses the challenge of transferring cognitive workload estimation models from controlled laboratory settings to realistic, non-stationary environments using Electroencephalography (EEG). A primary obstacle in this field is that standard classification methods often rely on discriminative artifacts, such as eye and muscle activity, which are highly condition-specific and fail to generalize across different tasks. The authors investigate whether adaptive spatial filtering can extract robust neural components that allow for successful classifier transfer between distinct experimental conditions. The research utilized EEG data from nine professional tugboat captains performing tasks in a realistic ship simulator. Two conditions were employed to induce varying levels of cognitive workload: a maneuvering task involving complex ship towing under difficult weather conditions, and an auditory n-back secondary task performed during simple sailing. The study compared three spatial filtering approaches: no filtering, Common Spatial Patterns (CSP), and a novel adaptive beamforming (BF) method. The BF approach used a head model to optimize filters for neural source extraction and continuously adapted to changing signal statistics via unsupervised updates. Features were derived from theta (4–7 Hz) and alpha (8–13 Hz) band variances and classified using Linear Discriminant Analysis. Results demonstrated that while no spatial filtering achieved the lowest classification error within individual conditions (10% for n-back, 22% for maneuvering), it failed completely in transfer scenarios, performing at chance level (45–53%). CSP also struggled with transfer, performing significantly worse than chance in one direction. In contrast, the adaptive beamforming approach was the only method capable of successfully classifying workload levels across both transfer directions, achieving normalized losses of 34% and 35%, which were significantly better than chance. Analysis of scalp patterns revealed that methods relying on low within-condition errors primarily detected eye and muscle artifacts. The adaptive beamforming filters, however, identified neural components with central scalp distributions, indicating that the successful transfer was due to the extraction of genuine neural activity rather than condition-specific artifacts. The study concludes that adaptive spatial filtering is essential for developing generalizable cognitive workload estimators. By optimizing filters to extract neural components and adapting to non-stationarities, the proposed beamforming approach overcomes the limitations of static methods like CSP, which often overfit to transient artifacts. This finding implies that for real-world applications involving changing tasks or environments, spatial filters must be dynamically adapted to maintain performance, ensuring that classification relies on stable neural markers rather than variable physiological noise.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• workload measurement