Global Flight Delay Patterns and Route Performance: A Business Intelligence and Predictive Analytics Approach Using Power BI

Abstract

Delays in flights are costly to the world aviation in terms of operations and economy. This paper will analyze flight delay and route performance based on a sample of 10,000 international flight records, 10 airlines, 10 routes, 6 geographic areas, 4 seasons, and 5 weather conditions. It utilized a systematized quantitative approach: the data were ready in Microsoft Power BI with Power Query and Data Analysis Expressions (DAX), and four interrelated interactive dashboards were built to facilitate geographic, temporal, airline, and predictive analysis. Six hypotheses were formally based on the existing theory and empirical literature and tested with one-way ANOVA with Tukey HSD post-hoc tests, Pearson correlation, chi-square tests, independent-samples t-tests, multiple linear regression, and feature importance analysis through Random Forest. The most prevalent structural predictor of delay severity was weather condition (ANOVA F (4,9995) = 3,213.26, p < 0.001; Storm vs Clear Cohen d = 2.38) and a discrete threshold effect delineating severe precipitation and all benign conditions. Seasonality resulted in two-tier structure where Summer and Winter delays were significantly higher than Spring and Autumn (ANOVA F (3,9996) = 107.42, p < 0.001; d ? 0.39). The delay was strongly and negatively correlated to route efficiency score (r = -0.958, p < 0.001), and statistically significant inter-airline differences proved the hypothesis that operational practice is an independently controllable source of performance variation (p = 0.001). Weather was verified to be the most significant predictive feature with a Random Forest model (importance = 0.578; R 2 = 0.94). The integrated BI framework incorporates these validated results into interactive dashboards, showing a template of decision support that can be replicated to real-world aviation operations management.

Introduction

Flight delays are a major operational and financial challenge for the global aviation industry, causing increased fuel consumption, crew scheduling disruptions, passenger compensation costs, and cascading delays across airline networks. Existing research has applied operations research, statistics, and machine learning to predict delays, but few studies combine statistical hypothesis testing with interactive business intelligence (BI) visualization. This study addresses that gap by analyzing a dataset of 10,000 international flights using inferential statistics and Microsoft Power BI dashboards to identify delay patterns, evaluate operational performance, and develop predictive analytics for flight delay management.

The study is grounded in five theoretical perspectives. Queuing Theory and Network Propagation explain how congestion at major airports causes cascading delays throughout airline networks. Attribution Theory distinguishes between controllable operational delays and uncontrollable weather-related delays, suggesting airlines differ in operational efficiency. The Resource-Based View (RBV) argues that airlines possess different operational capabilities that influence delay performance. The Theory of Constraints (TOC) identifies heavily utilized airports and routes as bottlenecks where targeted improvements can significantly reduce delays. Decision Support System (DSS) Theory provides the foundation for using Power BI dashboards to transform complex operational data into interactive decision-support tools.

The literature review identifies six major determinants of flight delays. Weather is consistently reported as the strongest predictor, with severe conditions such as storms and snow producing significantly larger delays than clear weather or rain. Seasonal demand also influences delays, with summer and winter experiencing the greatest congestion. Route efficiency is negatively associated with delays, indicating that efficient routes experience fewer disruptions. Airline operational practices contribute to significant differences in delay performance even when operating similar routes under similar weather conditions. Weather not only increases average delay but also shifts flights toward more severe delay categories. Finally, hybrid predictive approaches combining statistical regression and machine learning provide both interpretability and high predictive performance.

Based on the literature, the study formulates six hypotheses. H1 examines whether severe weather significantly increases average delays. H2 evaluates seasonal differences in delays. H3 investigates the relationship between route efficiency and delay duration. H4 tests whether airlines exhibit statistically significant differences in operational performance. H5 examines whether weather conditions influence the distribution of delay severity. H6 evaluates whether an integrated predictive model combining weather, season, airline, route, and operational variables can accurately predict flight delays.

The dataset consists of 10,000 simulated international flight records containing 25 variables across six categories, including scheduling information, delay metrics, route characteristics, airline identity, weather conditions, and fleet information. The data represent ten major international airlines, ten global routes, four seasons, five weather conditions, and four aircraft types. Data preprocessing in Microsoft Power BI includes duplicate removal, date-time standardization, validation of delay values, and creation of derived variables such as Delay Category, Peak Hour Indicator, and Delay Frequency using Power Query and DAX.

The statistical methodology aligns each hypothesis with an appropriate inferential technique. One-way ANOVA with Tukey HSD and Cohen's d evaluates weather and seasonal effects. Pearson correlation and independent t-tests assess the relationship between route efficiency and delays. Airline performance differences are analyzed using ANOVA and pairwise t-tests. Chi-square tests examine associations between weather conditions and delay severity categories. Multiple linear regression and Random Forest feature importance are used to develop an integrated predictive model that combines operational and environmental variables for flight delay prediction.

Conclusion

This paper has explored patterns of flight delays and route performance across the globe using a formally hypothesized multi-method empirical framework, combining four theoretical frameworks, namely, queuing theory, attribution theory, resource-based view and the theory of constraints, and a tight set of inferential statistical tests which have been executed in an interactive Power BI environment. Six hypotheses derived from the literature were all supported by the evidence. The structural determinant of delay of weather is the dominant structural determinant of delay, acting along a discrete threshold effect as opposed to a continuous severity gradient (H1), and the effect of weather is distributional, as well as a mean-level, restructuring the overall delay severity distribution to the high-cost tail in severe weather (H5). Seasonality creates a two-level design where Summer and Winter are statistically equal risky seasons (H2). Delay risk is almost perfectly represented by route efficiency (H3). There are statistically significant delay differences on routes shared by airlines (H4). In the Random Forest modelling, 94% of the delay variance is explained by an integrated model with weather as the leading predictor (H6). The main methodological addition would be the combination of formally specified, literature-based hypotheses with high-quality inferential testing in an interactive BI system - combining the traditionally distinct approaches of machine learning delay prediction, which is more concerned with accuracy, and operational BI reporting, which is more concerned with interpretability and decision-support value. The ensuing framework is reproducible, open and directly scalable to actual operational information within airline management systems, air navigation service providers or aviation regulatory databases. This framework should be tested in future studies against actual operational data, more fine-grained meteorological measurements (e.g., crosswind speed, visibility distance), longitudinal modelling should be used to test the temporal consistency of the identified seasonal and meteorological patterns, and cross-validated ensemble evaluation should be used to generate conservative accuracy estimates.

References

[1] AlBassam, S. A. A., & AlShahrani, D. N. (2025). Flight delay prediction: Evaluating machine learning algorithms for enhanced accuracy. PLOS ONE, 20(12), Article e0335141. https://doi.org/10.1371/journal.pone.0335141 [2] Alfarhood, M., Alotaibi, R., Abdulrahim, B., Einieh, A., Almousa, M., & Alkhanifer, A. (2024). Predicting flight delays with machine learning: A case study from Saudi Arabian Airlines. International Journal of Aerospace Engineering, 2024, Article 3385463. https://doi.org/10.1155/2024/3385463 [3] Bamford, D., & Xystouri, T. (2005). A case study of service failure and recovery within an international airline. Managing Service Quality, 15(3), 306–322. https://doi.org/10.1108/09604520510597814 [4] Barnhart, C., Belobaba, P., & Odoni, A. R. (2003). Applications of operations research in the air transport industry. Transportation Science, 37(4), 368–391. https://doi.org/10.1287/trsc.37.4.368.23276 [5] Barney, J. (1991). Firm resources and sustained competitive advantage. Journal of Management, 17(1), 99–120. https://doi.org/10.1177/014920639101700108 [6] Barros, C. P., & Wanke, P. (2015). An analysis of African airlines efficiency with two-stage TOPSIS and neural networks. Journal of Air Transport Management, 44–45, 90–102. https://doi.org/10.1016/j.jairtraman.2015.03.002 [7] Bisandu, D. B., & Moulitsas, I. (2023). A deep BiLSTM machine learning method for flight delay prediction classification. Journal of Aviation/Aerospace Education & Research, 32(2), 1–32. https://doi.org/10.58940/2329-258X.1992 [8] Cheng, B. L., Gan, C. C., Imrie, B. C., & Penman, S. (2008). Service recovery, customer satisfaction and customer loyalty: Evidence from Malaysia\'s hotel industry. International Journal of Hospitality Management, 3(1), 90–106. [9] Cook, A., & Tanner, G. (2015). European airline delay cost reference values: Updated and extended values (Version 4.1). Performance Review Unit, Eurocontrol. [10] Cumming, G. (2014). The new statistics: Why and how. Psychological Science, 25(1), 7–29. https://doi.org/10.1177/0956797613504966 [11] Dai, M. (2024). A hybrid machine learning-based model for predicting flight delay through aviation big data. Scientific Reports, 14, Article 4603. https://doi.org/10.1038/s41598-024-55217-z [12] Eckerson, W. W. (2010). Performance dashboards: Measuring, monitoring, and managing your business (2nd ed.). John Wiley & Sons. [13] Erlang, A. K. (1909). The theory of probabilities and telephone conversations. Nyt Tidsskrift for Matematik, 20(B), 33–39. [14] Fleurquin, P., Ramasco, J. J., & Eguiluz, V. M. (2013). Systemic delay propagation in the US airport network. Scientific Reports, 3, Article 1159. https://doi.org/10.1038/srep01159 [15] Goldratt, E. M. (1984). The goal: A process of ongoing improvement. North River Press. [16] Güvercin, M., Ferhatosmano?lu, N., & Gedik, B. (2021). Forecasting flight delays using clustered models based on airport networks. IEEE Transactions on Intelligent Transportation Systems, 22(5), 3179–3189. https://doi.org/10.1109/TITS.2020.2990960 [17] Hat?po?lu, ?., & Tosun, Ö. (2024). Predictive modeling of flight delays at an airport using machine learning methods. Applied Sciences, 14(13), Article 5472. https://doi.org/10.3390/app14135472 [18] Hat?po?lu, ?., Tosun, Ö., & Tosun, N. (2022). Flight delay prediction based with machine learning. LogForum, 18(1), 97–107. https://doi.org/10.17270/J.LOG.2022.655 [19] Heider, F. (1958). The psychology of interpersonal relations. John Wiley & Sons. [20] Hopp, W. J., & Spearman, M. L. (2001). Factory physics: Foundations of manufacturing management (2nd ed.). Irwin/McGraw-Hill. [21] International Air Transport Association. (2024). Annual review 2024. https://www.iata.org/contentassets/c81222d96c9a4e0bb4ff6ced0126f0bb/iata-annual-review-2024.pdf [22] Keen, P. G. W., & Scott Morton, M. S. (1978). Decision support systems: An organizational perspective. Addison-Wesley. [23] Li, Q., Guan, X., & Liu, J. (2023). A CNN-LSTM framework for flight delay prediction. Expert Systems with Applications, 227, Article 120287. https://doi.org/10.1016/j.eswa.2023.120287 [24] Li, Q., & Jing, R. (2022). Flight delay prediction from spatial and temporal perspective. Expert Systems with Applications, 205, Article 117662. https://doi.org/10.1016/j.eswa.2022.117662 [25] Mayer, C., & Sinai, T. (2003). Network effects, congestion externalities, and air traffic delays: Or why not all delays are evil. American Economic Review, 93(4), 1194–1215. https://doi.org/10.1257/000282803769206269 [26] Merkert, R., & Morrell, P. S. (2012). Mergers and acquisitions in aviation: Management and economic perspectives on the size of airlines. Transportation Research Part E: Logistics and Transportation Review, 48(4), 853–862. https://doi.org/10.1016/j.tre.2012.02.002 [27] Nguyen, C.-V. (2024). Air transport resilience, tourism and its impact on economic growth. Economies, 12(9), Article 236. https://doi.org/10.3390/economies12090236 [28] Penrose, E. T. (1959). The theory of the growth of the firm. Oxford University Press. [29] Pyrgiotis, N., Malone, K. M., & Odoni, A. (2013). Modelling delay propagation within an airport network. Transportation Research Part C: Emerging Technologies, 27, 60–75. https://doi.org/10.1016/j.trc.2011.09.003 [30] Qu, J., Wu, S., & Zhang, J. (2023). Flight delay propagation prediction based on deep learning. Mathematics, 11(3), Article 494. https://doi.org/10.3390/math11030494 [31] Rupp, N. G., & Holmes, G. M. (2006). An investigation into airline on-time performance. Applied Economics Letters, 13(3), 163–166. https://doi.org/10.1080/13504850500392635 [32] Sharda, R., Delen, D., & Turban, E. (2020). Analytics, data science, & artificial intelligence: Systems for decision support (11th ed.). Pearson. [33] Simon, H. A. (1960). The new science of management decision. Harper & Row. [34] Song, C., Ma, X., Ardizzone, C., & Zhuang, J. (2024). The adverse impact of flight delays on passenger satisfaction: An innovative prediction model utilizing wide & deep learning. Journal of Air Transport Management, 114, Article 102511. https://doi.org/10.1016/j.jairtraman.2023.102511 [35] Wandelt, S., Chen, X., & Sun, X. (2025). Flight delay prediction: A dissecting review of recent studies using machine learning. IEEE Transactions on Intelligent Transportation Systems, 26(4), 4283–4297. https://doi.org/10.1109/TITS.2025.3528536 [36] Wang, F., Bi, J., Xie, D., & Zhao, X. (2022). Flight delay forecasting and analysis of direct and indirect factors. IET Intelligent Transport Systems, 16(7), 890–907. https://doi.org/10.1049/itr2.12183 [37] Weiner, B. (1985). An attributional theory of achievement motivation and emotion. Psychological Review, 92(4), 548–573. https://doi.org/10.1037/0033-295X.92.4.548.

Copyright

Copyright © 2026 Dr. Anju Choudhary. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.