TY - GEN
T1 - ATC-WSA
T2 - 2nd CAAI International Conference on Artificial Intelligence, CICAI 2022
AU - Liu, Bo
AU - Wang, Xuanqian
AU - Dong, Jingjin
AU - Li, Di
AU - Lu, Feng
N1 - Publisher Copyright:
© 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.
PY - 2022
Y1 - 2022
N2 - Air traffic controllers (ATCs) are required to focus on flight information, make instant decisions and give instructions to pilots with high attention and responsibility. Human factors related to aviation risks should be monitored, such as fatigue, distraction, and so on. However, existing methods have two major problems: 1) Wearable or invasive devices may interfere with ATCs’ work; 2) Appropriate state indicator for ATCs is still not clear. Therefore, we propose a working state analysis solution, called ATC-WSA, and solve the above questions by 1) Computer vision and speech techniques without contact; 2) Specific models and indexes optimized by collected real ATCs’ data, including video, audio, annotation, and questionnaire. Three layers’ architecture is designed for AI detection, state analysis, and high-level indexes calculation. Overall, our demo can monitor and analyze the working state of ATCs and detect abnormal states in time. Key parts of this demo have already been applied to North China Air Traffic Control Center (Beijing) and the control tower of Beijing Capital International Airport.
AB - Air traffic controllers (ATCs) are required to focus on flight information, make instant decisions and give instructions to pilots with high attention and responsibility. Human factors related to aviation risks should be monitored, such as fatigue, distraction, and so on. However, existing methods have two major problems: 1) Wearable or invasive devices may interfere with ATCs’ work; 2) Appropriate state indicator for ATCs is still not clear. Therefore, we propose a working state analysis solution, called ATC-WSA, and solve the above questions by 1) Computer vision and speech techniques without contact; 2) Specific models and indexes optimized by collected real ATCs’ data, including video, audio, annotation, and questionnaire. Three layers’ architecture is designed for AI detection, state analysis, and high-level indexes calculation. Overall, our demo can monitor and analyze the working state of ATCs and detect abnormal states in time. Key parts of this demo have already been applied to North China Air Traffic Control Center (Beijing) and the control tower of Beijing Capital International Airport.
KW - Air traffic controller
KW - Keyword spotting
KW - State analysis
UR - https://www.scopus.com/pages/publications/85145263107
U2 - 10.1007/978-3-031-20503-3_42
DO - 10.1007/978-3-031-20503-3_42
M3 - 会议稿件
AN - SCOPUS:85145263107
SN - 9783031205026
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 521
EP - 525
BT - Artificial Intelligence - Second CAAI International Conference, CICAI 2022, Revised Selected Papers
A2 - Fang, Lu
A2 - Povey, Daniel
A2 - Zhai, Guangtao
A2 - Mei, Tao
A2 - Wang, Ruiping
PB - Springer Science and Business Media Deutschland GmbH
Y2 - 27 August 2022 through 28 August 2022
ER -