Real-time prediction of short-timescale fluctuations in cognitive workload
DOI: 10.1186/s41235-021-00289-y
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 implementing adaptive automation systems that can dynamically adjust to operators' cognitive workload in real time. While adaptive automation aims to mitigate performance issues caused by workload fluctuations (such as overload or neglect), current monitoring methods are often obtrusive or lack the temporal resolution required for second-by-second adjustments. The authors investigate whether short-timescale fluctuations in cognitive workload can be predicted using non-intrusive measures derived from primary task performance, thereby eliminating the need for continuous, intrusive workload monitoring during operation. To test this, the researchers conducted an experiment with 46 participants who managed a simulated fleet of unmanned aerial vehicles (UAVs) over two one-hour sessions. The primary task required participants to monitor UAV fuel levels and refuel them to prevent crashes, with difficulty manipulated by varying the number of UAVs (3, 5, or 7). Cognitive workload was measured using the Detection Response Task (DRT), an ISO-standard secondary task involving tactile stimuli presented every 3–5 seconds. The study first validated the DRT as a sensitive workload measure, confirming that reaction times and omission rates increased significantly with higher task difficulty. The core analysis employed between-subjects cross-validation to assess whether specific primary task events and situational awareness metrics could predict DRT performance (reaction time and omissions) in the immediate future. The results demonstrated that cognitive workload fluctuates rapidly in response to recent task events. Simple measures, such as the mere occurrence of task events, showed weak predictive ability. However, composite measures that indexed the operator’s situational awareness regarding UAV fuel levels were significantly more effective predictors of current workload. The predictive models, trained on subsets of participants, successfully generalized to out-of-sample data, indicating that real-time prediction of workload is feasible. The study confirms that spare cognitive capacity varies as a function of immediate task demands rather than just long-term averages. These findings imply that adaptive automation systems can be designed to predict operator workload without relying on continuous physiological monitoring or intrusive secondary tasks during operation. By leveraging predictive models based on primary task data collected during an initial training phase, automation can anticipate peaks and troughs in cognitive load. This approach offers a viable pathway for enhancing human-automation teaming in complex, high-pressure environments, such as fleet management, by allowing systems to provide assistance precisely when operators are most cognitively constrained.
Key finding
Cognitive workload, measured by the ISO 17488 DRT, varies rapidly (over seconds) as a function of recent primary-task events, and composite predictors capturing operators' situational awareness about fuel levels predict DRT RT and omissions in held-out data far better than predictors based only on immediately preceding task-event occurrences, supporting real-time predictive workload models for adaptive automation.
Methodology
other
Sample size: N=46 adults (University of Utah participant pool); two one-hour sessions per participant on different days.
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. Discovered via author_sweep_intake on 2026-05-27 (6 acquisition events logged).
| Stage | Outcome | Tool | Model | Prompt | Attempts | Completed |
|---|---|---|---|---|---|---|
| discover | success | author_sweep | — | — | 5 | 2026-05-28 |
| archive | success | unpaywall | — | — | 2 | 2026-06-02 |
| extract | success | cached | — | — | 2 | 2026-06-10 |
| clean | success | — | — | — | 1 | 2026-06-01 |
| chunk | success | — | — | — | 1 | 2026-06-01 |
| embed | success | — | — | — | 1 | 2026-06-02 |
| enrich | success | crossref | — | — | 1 | 2026-06-04 |
| promote | success | — | — | — | 2 | 2026-06-06 |
| summarize | success | llm | qwen3.6-27b-prismaquant | summ-v5 | 2 | 2026-06-10 |
| tag | success | vector_similarity | — | — | 17 | 2026-06-11 |
| verify | success | — | — | — | 2 | 2026-06-10 |
Summary generated by qwen3.6-27b-prismaquant on 2026-06-10; verification: verified.
Topics
Ranked by relevance to this paper. Hover a topic for its definition.
Information type
What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).
- Empirical Findings: physiological data
- Theoretical Contribution: theory or model, computational model