Portal de Conferências da UFSC, XX Sitraer

Tamanho da fonte: 
Explainable Artificial Intelligence applied to active runway configuration and dynamic airport capacity prediction
Igor Galhano Gomes, Mayara Condé Rocha Murça, Marcelo Xavier Guterres

Última alteração: 2023-09-26


Despite the widespread use of machine learning methods in a variety of studies in the aviation industry in recent years, there has been limited progress in achieving explainability for the results generated by these models. Given the importance of explainable predictive models for trustworthy decision support in the context of Air Traffic Management (ATM), this study explores the use of Multilayer Perceptron and Random Forests, along with a Local Interpretable Model-agnostic Explanations (LIME) approach, to develop predictive models capable of generating interpretable forecasts of airport runway configuration and dynamic capacity. The machine learning process is based on historical weather, flight demand and actual air traffic movement data for Sao Paulo/Guarulhos International Airport. The predictive models are able to deliver runway configuration forecasts with an accuracy higher than 87.73% and arrival/departure capacity forecasts with errors as low as 1.309. With LIME, consistent explanations are obtained through the models on each prediction. An interactive predictive tool is created to output the runway configuration and capacity forecasts as well as the main contributing factors for each forecast within a 24-hour forecasting horizon. This research aims to improve decision-making by integrating predictive models and explanatory models into applications that provide useful information for ATM.


Avery, J., & Balakrishnan, H. (2015). Predicting Airport Runway Configuration: A Discrete-Choice Modeling Approach. Eleventh USA/Europe Air Traffic Management Research and Development Seminar, (pp. 1–11). Lisbon.

Choi, S., & Kim, Y. J. (2021). Artificial neural network models for airport capacity prediction. Journal of Air Transport Management, 97, 102146. doi:https://doi.org/10.1016/j.jairtraman.2021.102146

DECEA (2015). Portaria DECEA n° 78/DGCEA, de 23 de março de 2015. Aprova a reedição do MCA 100-14, Manual que trata da 'Capacidade do Sistema de Pistas'. Boletim do Comando da Aeronautica.

Degas, A., Islam, M. R., Hurter, C., Barua, S., Rahman, H., Poudel, M., Ruscio, D., Ahmed, M. U., Begum, S., Rahman, M. A., Bonelli, S., Cartocci, G., Flumeri, G. D., Borghini, G., Babiloni, F., and Arico, P. (2022). A Survey on Artificial Intelligence (AI) and eXplainable AI in Air Traffic Management: Current Trends and Development with Future Research Trajectory. Applied Sciences, 12. doi:10.3390/app12031295

Jiao, Q.-Y., Li, N., Zheng, Z.-G., Feng, Z., Ren, G.-S., & Qiang, Y.-G. (2023). Attention Mechanism-Based Deep Learning Method for Predicting Airport Acceptance Rate: A Case Study of Hong Kong Airport. Journal of Aeronautics, Astronautics and Aviation, 55, 185–200.

Rebollo, J., Khater, S., & Coupe, W. J. (2021). A Recursive Multi-step Machine Learning Approach for Airport Configuration Prediction. AIAA AVIATION 2021 FORUM, (p. 2406). doi:10.2514/6.2021-2406

Ribeiro, M. T., Singh, S., & Guestrin, C. (2016). "Why Should I Trust You?": Explaining the Predictions of Any Classifier. Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (pp. 1135–-1144). New York, NY, USA: Association for Computing Machinery. doi:10.1145/2939672.2939778

Wang, Y., & Zhang, Y. (2021). Prediction of runway configurations and airport acceptance rates for multi-airport system using gridded weather forecast. Transportation Research Part C: Emerging Technologies, 125, 103049. doi:https://doi.org/10.1016/j.trc.2021.103049

Woo, C. J., Goh, S. K., Alam, S., Ferdaus, M. M., & Ellejmi, M. (2022). A runway exit prediction model with visually explainable machine decisions. 2022 International Conference on Research in Air Transportation (ICRAT 2022), (pp. 1–9).

Um cadastro no sistema é obrigatório para visualizar os documentos. Clique aqui para criar um cadastro.