TY - GEN
T1 - Estimation of a Pilot’s Workload In-Flight Using External Fluctuation Factors
T2 - 21st Congress of the International Ergonomics Association, IEA 2021
AU - Mekata, Yuki
AU - Shiina, Kenta
AU - Osawa, Ayumu
AU - Nakanishi, Miwa
N1 - Publisher Copyright:
© 2021, The Author(s), under exclusive license to Springer Nature Switzerland AG.
PY - 2021
Y1 - 2021
N2 - Pilots must manage their workload correctly to achieve a safe operation. However, few studies have attempted to estimate a pilot’s workload in real-time, and there are no established methods of doing so. To provide more direct support for a pilot’s workload management, we attempted to construct a model that estimates the pilot’s future workload using data on various flight-related parameters that can be acquired in real-time. Participants conducted simulated flights using the flight simulator. Based on the data obtained from these simulations, we used machine learning to construct a model to estimate the workload level 30 s later on a five-point scale. This model correctly estimated 32.0% of the test data, and in 72.3% of the test data, the deviation between the subjective value and the estimated value was within one workload level. We implemented a system that presents the estimated workload level to the pilot in real-time, and from the review of a license holder who conducted simulated flights using the proposed system, we confirmed that the system is effective for workload management.
AB - Pilots must manage their workload correctly to achieve a safe operation. However, few studies have attempted to estimate a pilot’s workload in real-time, and there are no established methods of doing so. To provide more direct support for a pilot’s workload management, we attempted to construct a model that estimates the pilot’s future workload using data on various flight-related parameters that can be acquired in real-time. Participants conducted simulated flights using the flight simulator. Based on the data obtained from these simulations, we used machine learning to construct a model to estimate the workload level 30 s later on a five-point scale. This model correctly estimated 32.0% of the test data, and in 72.3% of the test data, the deviation between the subjective value and the estimated value was within one workload level. We implemented a system that presents the estimated workload level to the pilot in real-time, and from the review of a license holder who conducted simulated flights using the proposed system, we confirmed that the system is effective for workload management.
KW - Aircraft
KW - Flight data
KW - Machine learning
KW - Pilot’s workload level
KW - Workload management
UR - http://www.scopus.com/inward/record.url?scp=85111114706&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85111114706&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-74608-7_20
DO - 10.1007/978-3-030-74608-7_20
M3 - Conference contribution
AN - SCOPUS:85111114706
SN - 9783030746070
T3 - Lecture Notes in Networks and Systems
SP - 150
EP - 158
BT - Proceedings of the 21st Congress of the International Ergonomics Association, IEA 2021 - Sector Based Ergonomics
A2 - Black, Nancy L.
A2 - Neumann, W. Patrick
A2 - Noy, Ian
PB - Springer Science and Business Media Deutschland GmbH
Y2 - 13 June 2021 through 18 June 2021
ER -