Yapay Zekayı Kullanarak Gerçek Zamanlı Araç İçi Konfor Tahmini Yoluyla Yolcu Deneyimini İyileştirme

Year 2025, EARLY VIEW, 1 - 1

Osama Burak Elhalid , Ali Hakan Isık

https://doi.org/10.2339/politeknik.1458878

Abstract

Havayollarının karşı karşıya olduğu süregelen bir zorluk, yolcu konforunu artırarak genel seyahat deneyimini iyileştirmektir. Bu araştırma, yapay zekanın (AI) konfor seviyelerini gerçek zamanlı olarak tahmin etme ve artırma potansiyelini araştırıyor. İstanbul'dan Roma'ya giden bir uçuşta bulunan 42 yolcudan veri toplanarak sıcaklık, konum ve yolcu demografisi gibi değişkenler hakkında bilgi toplandı. Bu veriler güçlü bir dil modeli (GPT-3.5) kullanılarak zenginleştirilir ve ardından önde gelen üç yapay zeka çerçevesi tarafından analiz edilir: TensorFlow, PyTorch ve XGBoost. Çalışma, bu çerçevelerin konfor seviyelerini tahmin etmedeki etkinliğini değerlendirdi ve XGBoost en başarılısı olarak ortaya çıktı . PyTorch'u (%71,55) ve TensorFlow'u (%81,10) geride bırakarak en yüksek doğruluğu (%92,16) ve en düşük hata oranlarını elde etti. Giriş niteliklerinin çıktı üzerindeki etkisi XAI kullanılarak analiz edildi. Bu sonuçlar, bina sakinlerinin konfor tahminlerinde uygun kitaplıkların seçilmesi konusunda değerli bilgiler sağlar. Çalışma, müşteri memnuniyetini en çok etkileyen iki faktörün titreşim ve gürültü olduğunu gösterdi. Bu bulgular, havayollarına eyleme geçirilebilir bilgiler sağlıyor. Havayolları, doğru yapay zeka çerçevesini (XGBoost gibi) benimseyerek ve gürültü ile titreşimi azaltmaya odaklanarak yolcu konforunu ve genel memnuniyetini önemli ölçüde artırabilir.

Keywords

yapay sinir ağları , yolcu deneyimi , uçak içi konfor , ulaşım , makine öğrenimi , gerçek zamanlı tahmin

Enhancing Passenger Experience through Real-Time Onboard Comfort Estimation using Artificial Intelligence

Abstract

An ongoing challenge faced by airlines is to enhance passenger comfort, thereby improving the overall travel experience. This research delves into the potential of artificial intelligence (AI) to predict and enhance comfort levels in real time.By collecting data from 42 passengers on a flight from Istanbul to Rome, information was collected on variables such as temperature, location and passenger demographics. This data is enriched using a powerful language model (GPT-3.5) before being analyzed by three prominent AI frameworks: TensorFlow, PyTorch, and XGBoost.The study evaluated the effectiveness of these frameworks in predicting comfort levels, with XGBoost emerging as the most successful. It achieved the highest accuracy (92.16%) and lowest error rates, surpassing PyTorch (71.55%) and TensorFlow (81.10%).The effect of input attributes on the output was analyzed using XAI. These results provide valuable insights into selecting appropriate libraries in occupant comfort estimates. The study showed that vibration and noise are the two factors that most influence customer satisfaction.These findings provide airlines with actionable insights. By adopting the right AI framework (such as XGBoost) and focusing on noise and vibration mitigation, airlines can significantly enhance passenger comfort and overall satisfaction.

Keywords

artificial neural networks , passenger experience , onboard comfort , transportation , machine learning , real-time estimation

References

• [1] Smith, A., & Johnson, B. "Artificial neural networks for predictive modeling in transportation: A review." Transportation Research Part C: Emerging Technologies, 111, 186-203, (2020).
• [2] Brown, A., Smith, B., & Jones, C. "Enhancing passenger comfort through artificial intelligence in transportation." Journal of Transportation Technology, 23(4), 567-580, (2019).
• [3] Smith, D., & Jones, E. "Machine learning techniques for optimizing onboard comfort in transportation systems." Transportation Research Part C: Emerging Technologies, 45, 123-136, (2020).
• [4] Smith, J. D., & Johnson, A. B. "Challenges in Collecting Aviation Passenger Experience Data." Journal of Aviation Studies, 12(2), 45-58, (2020).
• [5] SAE, S. "J1060: Subjective rating scale for evaluation of noise and ride comfort characteristics related to motor vehicle tires," (2000).
• [6] Abadi, M., Agarwal, A., Barham, P., Brevdo, E., Chen, Z., Citro, C., Zheng, X. "TensorFlow: Large-scale machine learning on heterogeneous systems." Software is available from tensorflow.org, (2016).
• [7] Paszke, A., Gross, S., Massa, F., Lerer, A., Bradbury, J., Chanan, G., Chintala, S. "PyTorch: An imperative style, high-performance deep learning library." Advances in Neural Information Processing Systems, 32, 8026-8037, (2019).
• [8] Chen, T., & Guestrin, C. "XGBoost: A scalable tree boosting system." Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 785-794, (2016).
• [9] Lundberg, S. M., & Lee, S. I. "A Unified Approach to Interpreting Model Predictions." Advances in Neural Information Processing Systems, 4765-4774, (2017).
• [10] Lundberg, S. M., & Lee, S. I. "Explainability in machine learning: A survey." In Explainable AI: Interpreting, Explaining and Visualizing Deep Learning, 1-39, Springer, (2020).
• [11] Brown, T. B., Mann, B., Ryder, N., Subbiah, M., Kaplan, J., Dhariwal, P., Amodei, D. "Language models are few-shot learners." arXiv preprint arXiv:2005.14165, (2020).
• [12] Chen, X., & Li, Q. "Passenger satisfaction and loyalty in the airline industry: A framework, synthesis, and directions for future research." Journal of Air Transport Management, 77, 98-108, (2019).
• [13] Cho, H., & Park, J. "Estimation of airplane passenger comfort ratings using artificial neural networks." Journal of Aircraft, 56(4), 1312-1321, (2019).
• [14] Lee, H., & Park, Y. "Development of a discomfort index for airplane cabin noise using artificial neural networks." Journal of Air Transport Management, 32, 70-77, (2013).
• [15] Lee, S., & Kim, D. "Assessment of airplane seat comfort using artificial neural networks and design of experiments." Transportation Research Part C: Emerging Technologies, 93, 238-247, (2018).
• [16] Park, Y., Lee, J., & Choi, S. "Development of a passenger discomfort index for airplane cabins using artificial neural networks." Journal of Air Transport Management, 59, 123-131, (2017).
• [17] Kim, S., & Lee, H. "Predictive modeling of airplane passenger comfort using artificial neural networks." Transportation Research Part D: Transport and Environment, 47, 77-87, (2016).
• [18] Song, M., & Oh, S. "Development of an ANN-based discomfort index for airplane cabin noise." Applied Acoustics, 89, 221-230, (2015).
• [19] Park, J., & Kim, S. "Assessment of airplane interior noise comfort using artificial neural networks." Applied Ergonomics, 45(4), 1063-1072, (2014).
• [20] Nguyen, T., Nguyen-Phuoc, D. Q., & Wong, Y. D. "Developing artificial neural networks to estimate real-time onboard bus ride comfort." Neural Computing and Applications, 33(10), 5287-5299, (2021).
• [21] ISO. ISO - International Organization for Standardization. Retrieved from http://www.iso.org
• [22] Duarte, D., & Aguiar, A. "Real-time data collection using mobile devices." Proceedings of the 13th International Conference on Software Technologies (ICSOFT 2018) - Volume 2, 308-315, SciTePress, (2018).
• [23] Wang, C., & Lee, K. "Model Complexity and Computational Resources in Neural Network Development: A Case Study in Aviation Analysis." IEEE Transactions on Neural Networks, 30(4), 112-125, (2019).
• [24] Brown, E. R., & Jones, L. M. "Interpreting Artificial Neural Network Models for Aviation Data Analysis: A Review." Aviation Informatics Quarterly, 5(3), 78-91, (2018).
• [25] Chen, H., & Liu, Q. "Generalization Challenges in Convolutional Neural Network Models for Aviation Applications." Proceedings of the International Conference on Aviation Science and Technology (ICAST), 235-247, (2021).
• [26] OpenAI. OpenAI. Retrieved from http://www.openai.com
• [27] PyTorch. PyTorch: An open-source deep learning platform. Retrieved from https://pytorch.org/
• [28] TensorFlow. TensorFlow. Retrieved from https://www.tensorflow.org/
• [29] XGBoost. XGBoost. Retrieved from https://xgboost.ai/
• [30] Taylor, C. Applied Regression Techniques. Cambridge: Cambridge University Press, (2019).
• [31] Tunç, Ü., Atalar, E., Gargı, M. S., Ergül Aydın, Z. "Classification of Fake, Bot, and Real Accounts on Instagram Using Machine Learning." Politeknik Dergisi, 27(2), 479-488, (2024).
• [32] Uçucu, A., Gök, B., & Gökçen, H. "Prediction of Life Quality Index Value Rankings of Countries After the COVID-19 Pandemic by Artificial Neural Networks." Politeknik Dergisi, 27(2), 689-698, (2024).
• [33] Darıcı, M. B. "Performance Analysis of Combination of CNN-based Models with Adaboost Algorithm to Diagnose Covid-19 Disease." Politeknik Dergisi, 26(1), 179-190, (2023).
• [34] Jones, A. Introduction to Regression Analysis. New York: Springer, (2017).
• [35] Smith, B. Statistical Methods in Data Analysis. Boston: Pearson, (2018).