Research on pilot perceived intention method based on EEG and visual fixation fusion
DOI: 10.21203/rs.3.rs-1537774/v1
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 accurately and rapidly assessing pilot situational awareness (SA), a critical factor in aviation safety. Traditional methods relying solely on visual gaze analysis suffer from high latency and significant deviation due to the complexity of pilot cognitive states. To overcome these limitations, the authors propose a novel method that fuses electroencephalogram (EEG) data with visual fixation and flight control inputs. The research aims to develop a real-time, high-precision classification model for pilot perceived intention, specifically focusing on altitude, airspeed, attitude, and heading during simulated flight tasks. The methodology involves simulated flight experiments using an L39C trainer model on a motion platform. Flight cadets performed five-side flight maneuvers while researchers synchronously collected data from a non-wearable eye tracker, flight control inputs, and a 32-channel EEG cap. The study first established ground-truth labels for perceived intention events by analyzing flight parameters and visual gaze patterns. Visual data was processed using Region of Interest (AOI) analysis, extracting features such as gaze drift rate differential entropy and fixation duration. Flight control data was characterized by signal power energy. These multimodal features were used to label 1-second EEG segments corresponding to specific perceptual intentions. The core analytical model is a Transformer-based neural network designed to handle multi-dimensional sequential data. It utilizes a multi-head self-attention mechanism to extract spatiotemporal features from the EEG signals, followed by feature fusion through fully connected layers to classify the pilot’s perceived intention. The experimental results demonstrate that the proposed Transformer-based model achieves a classification accuracy of 96% on the constructed dataset. The system is capable of evaluating and classifying the pilot’s situational awareness state within a 5-second analysis window. The integration of EEG data with visual and control inputs effectively resolves the uncertainty associated with single-channel gaze analysis, allowing for more robust detection of pilot intent. The model successfully distinguishes between different types of perceived intentions, such as altitude maintenance versus heading correction, by correlating neural activity with specific instrument scanning behaviors and control adjustments. The significance of this research lies in its potential to enhance pilot training and selection processes by providing an objective, real-time assessment of situational awareness. By moving beyond post-experiment analysis to near-real-time detection, the method offers a practical tool for monitoring pilot cognitive load and performance during flight. The high accuracy and low latency of the model suggest its applicability in future aviation systems for improving flight safety and operational efficiency. This work contributes to the field of human-machine interaction by demonstrating the efficacy of deep learning architectures, specifically Transformers, in processing complex, multimodal physiological and behavioral data.
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.
| Stage | Outcome | Tool | Model | Prompt | Attempts | Completed |
|---|---|---|---|---|---|---|
| discover | success | Crossref | — | — | 1 | 2026-08-09 |
| archive | success | canonical_url | — | — | 1 | 2026-08-09 |
| extract | success | cached | — | — | 3 | 2026-08-10 |
| clean | success | clean | — | — | 1 | 2026-08-09 |
| chunk | success | chunk | — | — | 1 | 2026-08-09 |
| embed | success | embed | Qwen/Qwen3-Embedding-8B | — | 1 | 2026-08-09 |
| promote | success | — | — | — | 1 | 2026-08-09 |
| summarize | success | llm | qwen3.6-27b-nvidia | summ-v5 | 2 | 2026-08-10 |
| tag | success | vector_similarity | — | — | 11 | 2026-08-11 |
| verify | success | — | — | — | 2 | 2026-08-10 |
Summary generated by qwen3.6-27b-nvidia on 2026-08-10; 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
- Methodological Resource: tool software