Junghyun (Andy) Kim
(Advisor: Prof. Dimitri Mavris]

will propose a doctoral thesis entitled,

A Data-Driven Approach using Machine Learning for Real-Time Flight Path Optimization

On

Monday, April 6 at 2:00 p.m.
https://bluejeans.com/310489222

Abstract
As aviation traffic continues to grow, most airlines have growing concerns about flight delays, which are directly related to real costs for the airlines. Since most delays are caused by weather, pilots (or flight dispatchers) typically gather all available weather information prior to departure to create an efficient and safe flight plan. However, they may have to perform in-flight re-planning because weather information can significantly change after the original flight plan is created. The good news is that they are conscientious about their work and rarely cause accidents in the airspace. One potential issue, however, is that pilots today perform in-flight re-planning manually. The manual decision process may not be an issue because they generally make the right decisions regarding safety. However, the advent of new communication and technology will bring more information into the cockpit in the near future; therefore, they may be faced with information overload. This could prevent them from making quick and appropriate decisions especially if they must consider a large volume of information, leading to potential human error. Another potential issue is that weather forecasts being currently used in the aviation industry provide relatively unreliable information and are not accessible fast enough so that it challenges pilots to perform in-flight re-planning more accurately and frequently.

This research attempts to resolve the potential issues by performing in-flight re-planning automatically in a more accurate and frequent manner. More specifically, this research has three objectives: 1) Create a supervised machine learning model to obtain continuous wind information, 2) Perform short-term (i.e. every 10 minutes) convective weather predictions with an unsupervised machine learning model to obtain a more reliable and up-to-date area of convective weather, and 3) Perform A* based free-flight path optimization with a network that considers wind, convective weather, and the U.S. airspace infrastructure. This research concentrates specifically on a subset of the flight path optimization problem that pertains only to commercial en-route flights in the contiguous 48 states of the U.S. to accomplish the goals of reducing weather-related delays and improving fuel efficiency.

This study will establish not only an automated framework that enables real-time flight path optimization in a more accurate and frequent manner but also a basis for optimizing routes for all categories of airplanes such as commercial jet and business jet.

Committee

• Prof. Dimitri Mavris – School of Aerospace Engineering (advisor)
• Prof. Daniel Schrage – School of Aerospace Engineering
• Prof. Polo Chau – School of Computational Science and Engineering
• Prof. Chao Zhang – School of Computational Science and Engineering
• Dr. Simon Briceno – Jaunt Air Mobility

Meeting Information

• Meeting link:      https://bluejeans.com/310489222
• Meeting ID:         310 489 222
• Passcode:            3129