Data Study of Performance in Premium Airlines

Posted on Jul 19, 2017
The skills the author demoed here can be learned through taking Data Science with Machine Learning bootcamp with NYC Data Science Academy.

Data Study of Performance in Premium Airlines

Have you ever found that, the flight for which you paid a premium fare did not live up to your expectations? Maybe you would like to be able to select an airline that has a better track record. But where to find the data and information?

Introduction

As airlines transported 3.6 billion passengers in the year 2016, which is about 800 million more than in 2011 (IATA, 2016). Over the last 40 years, global air travel has increased tremendously as it has become more affordable. This decline in fare price is the response to the airline deregulation act from 1978 and to the dynamics in the elasticity of demand for air transportation. This decline in the fare price is desired by most passengers as research indicates that the ticket price is still the most important feature on which passengers choose an airline, despite other investments such as loyalty programs and the overall improvement of the customer’s experience (Skyscanner, 2009).  

However, booking the flight with the lowest fare is not always the best strategy. low cost carriers are often also leaders in late arrivals, low customer satisfaction, uncomfortable cabins and unsatisfactory frequent flyer programs (NBC News, 2017).

While one group of passengers aims to minimize flight fare, a second group has other priorities. For example, a flight is far more valuable for a salesperson who suddenly has an opportunity to visit an important client than to someone visiting a friend or relative abroad. Consequently, the salesperson is more willing to pay a higher fare in order to make the appointment (Airline Economics, 2017). As the price of the flight is of secondary concern, the salesperson is interested in other performance indicators for flight selection.

That raises the following research question: Can performance indicators on flight duration support passengers in the decision-making process of airline selection?

Data Set

To assess airline operation, a wide variety of indicators are available; however, not all of these indicators are relevant for the passenger. Relevant indicators of operational performance might be: percentage of flights delayed, average minutes delay per flight, percentage of flights delayed per reason, and for example the Number of flights to a particular destination (NWDS, 2017). In order to present such insights to passengers, a reliable longitudinal data set that allows for the analysis of multiple airlines over an extended period of time is required.

The time period should be chosen carefully, as airline performance tends to sway over seasons, which could sequentially influence the results (Bureau of Transport Statistics, 2002). Consequently, the time period is set to the last month of flight information in order to prevent seasonal bias and to ensure that passengers are informed about the current state of affairs. A dataset that allows for such an analysis is the U.S. Airline On-Time Performance dataset from the Bureau of Transportation Statistics, U.S. Department of Transportation (https://www.transtats.bts.gov/DL_SelectFields.asp?Table_ID=236&DB_Short_Name=On-Time) .

11 Data Sets Used for the Analysis

The USA Airline On-Time Performance dataset contains 479,000 observations on 12 airlines, observed over the period April 2017. Furthermore, the dataset contains a multitude of variables from which 11 were used for the analysis:

  1. Unique Carrier: A code representing the FAA designated airline code
  2. Origin: A code representing the airport from which the flight originated
  3. Destination: A code representing the airport to which the flight is flying
  4. Taxi In: The recorded taxi time toward the gate after wheels on time, recorded in minutes
  5. Taxi Out: The recorded taxi out time toward the runway till wheels of time, recorded in minutes
  6. Air Time: Flight time, recorded in minutes
  7. Carrier Delay: Delays caused by the carrier like aircraft cleaning, aircraft repairs, etc., recorded in minutes
  8. Weather Delay: Weather delay is caused by extreme or hazardous weather forecast that are or manifest themselves on point of departure, on-route, or on point of arrival, recorded in minutes
  9. NAS Delay: Delay that is within the control of the National Airspace System (NAS) may include: non-extreme weather conditions, airport operations, heavy traffic volume, air traffic control, etc., recorded in minutes
  10. Security Delay: Security delay is caused by evacuation of a terminal or concourse, re-boarding of aircraft because of security breach, inoperative screening equipment and/or long lines in excess of 29 minutes at screening areas, recorded in minutes.
  11. Late Aircraft Delay: Arrival delay at an airport due to the late arrival of the same aircraft at a previous airport, recorded in minutes.

Methodology

In order to inform passengers on airline performance indicators, an application was constructed using R and Shiny. The application can be found at the following internet location: https://stevenjongerden.shinyapps.io/flightstatusapp/.

To evaluate the functionality and usability of the shiny application, a flight scenario, consisting of a flight from Atlanta Airport to San Francisco Airport on the 16th of Aug, returning on the 23rd of August, was designed.. The possible flights options were determined by using Sky Scanner, as well as the pricing options for these flights. Sequentially, in order to determine the “best” flight, the presented performance indicators and statistics from the Shiny application were used.

Data Findings and Results on Airline Performance

Overall Airline Performance

For the trip from Atlanta to San Francisco, there are three possible options for direct flights: Delta Airlines, Frontier Airlines and United Airlines. These options were found using Sky Scanner and align with the information available in the Shiny application. For these airlines, the following airline wide information is available:

Table 1: Overall Airline Performance

Number of Airplane Number of Destination Percentage of flights delayed
Delta Airlines 145 143 11.18%
Frontier Airlines 52 49 19.29%
United Airlines 85 83 15.36%

From these results, one could infer that the performance of Delta Airlines with regards to delayed flights is better in comparison with Frontier Airlines and United Airlines. This might be a first initial indication of the airlines performance for the route from Atlanta to San Francisco; however, does not necessarily mean that Delta Airlines performs on average for that particular route. Consequently, route specific analysis should be performed.

Route Specific Airline Performance

The Shiny application also provides the functionality to compare specific routes, as route performance might deviate from overall performance. When the airlines are compared for the route Atlanta to San Francisco, the following route performance is presented:

Table 2: Route Specific Airline Performance

(min) Taxi Out Air Time Taxi In Duration Delay
Delta Airlines 20 +/- 1 279 +/-1 11 +/- 1 310 +/- 2 9 +/- 3
Frontier Airlines 22 +/- 4 286 +/- 5 24 +/- 6 332 +/- 10 17 +/- 10
United Airlines 16 +/- 1 282 +/- 2 8 +/- 1 306 +/- 2 11 +/- 7
Density plot to compare Airlines flights

Comparison of airlines for the flight from ATL to SFO

Observations

For this route, the presented route performance indicates that the flight duration (taxi out + air time + taxi in) is the longest for Frontier Airlines, and comparable for Delta Air Lines and United Airlines. It is also possible to infer that the overall delay for Frontier Airlines is longer than Delta Air Lines and United Flights, which is in line with the earlier findings from the average airline performance.

However, from this information, as Delta Airlines and United Airlines are somehow comparable, it is not yet possible to conclude which airline performs better. Consequently, the flights are compared using ANOVA and a Tuckey-HSD post-hoc test (a test to compare the average flight duration for more than two groups). The table below presents the results and indicates that Delta Air Lines flights and United Airlines flights are both statistically shorter than Frontier Airlines, but that there is no difference between United Airlines and Delta Air Lines.

Table 3:  Route Specific Airline Performance Comparison

Flight Duration Comparison Statistical Difference Flight Duration
Frontier / Delta Different Longer
United / Delta Not Different The Same
United / Frontier Different Shorter

Therefore, based on the average airline performance and specific route performance it is possible to conclude that Delta Airlines shows the best punctuality and flight duration, and therefore is the preferred airline for this particular route and time.

Flight Time Selection

At this point of the selection process, the date, airports and airline have been selected, which leaves the flight time. As Delta Air Lines provides seven opportunities on the 16th of Aug to fly to San Francisco. The passengers can select the flight time while taking into account factors that might influence the delay of the flight. Delay information for Atlanta International Airport indicates that as the day passes, more flights become delayed.

Consequently, you’re better off flying earlier in the day if you want to reduce the probability of a delayed departure This selection can be made without having to take into consideration any price changes as the flight fare is the same during the entire day.

Conclusion

Since the era of mass transportation, the price flights have become ever more important for commuters. However, another group of travelers exists that, next to the price of a flight, also highly ranks the punctuality of a flight. To facilitate this group of travelers in the flight selection process, a Shiny application was constructed using data from the Bureau of Transportation Statistics, U.S. Department of Transportation.

The results from a scenario study indicate that information on flight performance can assists in the selection of airlines for specific routes within the United States and assist in the selection for specific flight times in order to reduce the probability of delay. However, the findings presented within the document are limited to U.S Airlines for the period of April 2017 and therefore do not reflect future airline performance

Disclaimer

This publication has been prepared solely for illustration, educational and or discussion purposes. It does not constitute research of sufficient quality and under no circumstances should this publication or the information contained in them be used or considered for the selection of airlines.

Sources:

  1. AvJobs. (2017). Airline Economics. Retrieved from http://www.avjobs.com/history/airline-economics.asp
  2. Bureau of Transport Statistics. (2002). [A Time Series Analysis of Airline Delay]. Retrieved from https://www.rita.dot.gov/bts/sites/rita.dot.gov.bts/files/publications/transportation_indicators/
    january_2002/Special/html/A_Time_Series_Analysis_of_Airline_Delay.html
  3. Bureau of Transportation Statistics. (2017). On-Time Performance. Retrieved from https://www.transtats.bts.gov/DL_SelectFields.asp?Table_ID=236&DB_Short_Name=On-Time
  4. IATA. (2016). Airlines to Welcome 3.6 Billion Passengers in 2016. Retrieved from http://www.iata.org/pressroom/pr/Pages/2012-12-06-01.aspx
  5. NBC News. (2017). Why Booking the Cheapest Flight Isn’t Always the Best Strategy.
    Retrieved from http://www.nbcnews.com/business/travel/booking-cheapest-flight-it-could-cost-you-other-ways-n729701
  6. NWDS. (2017). Key Performance Indicators for Airlines and Airports. Retrieved from NWDS https://www.asms-pro.com/Modules/SafetyAssurance/ListofAirportKeyPerformanceIndicators.aspx
  7. Sky scanner. (2017). [Sky scanner]. Retrieved from https://www.skyscanner.net/press-releases/flight-price-and-convenience-most-important-flyers-say-skyscanner-users

About Author

Steven Jongerden

Steven graduated summa cum laude from the Delft University of Technology with a Masters degree in Engineering and Policy Analysis and a Bachelors degree in Aerospace Engineering. He is currently a Data Science Consultant employed by Capgemini Netherlands....
View all posts by Steven Jongerden >

Related Articles

Leave a Comment

Jack July 22, 2017
Thanks for sharing. Well-written and insightful.

View Posts by Categories


Our Recent Popular Posts


View Posts by Tags

#python #trainwithnycdsa 2019 2020 Revenue 3-points agriculture air quality airbnb airline alcohol Alex Baransky algorithm alumni Alumni Interview Alumni Reviews Alumni Spotlight alumni story Alumnus ames dataset ames housing dataset apartment rent API Application artist aws bank loans beautiful soup Best Bootcamp Best Data Science 2019 Best Data Science Bootcamp Best Data Science Bootcamp 2020 Best Ranked Big Data Book Launch Book-Signing bootcamp Bootcamp Alumni Bootcamp Prep boston safety Bundles cake recipe California Cancer Research capstone car price Career Career Day citibike classic cars classpass clustering Coding Course Demo Course Report covid 19 credit credit card crime frequency crops D3.js data data analysis Data Analyst data analytics data for tripadvisor reviews data science Data Science Academy Data Science Bootcamp Data science jobs Data Science Reviews Data Scientist Data Scientist Jobs data visualization database Deep Learning Demo Day Discount disney dplyr drug data e-commerce economy employee employee burnout employer networking environment feature engineering Finance Financial Data Science fitness studio Flask flight delay gbm Get Hired ggplot2 googleVis H20 Hadoop hallmark holiday movie happiness healthcare frauds higgs boson Hiring hiring partner events Hiring Partners hotels housing housing data housing predictions housing price hy-vee Income Industry Experts Injuries Instructor Blog Instructor Interview insurance italki Job Job Placement Jobs Jon Krohn JP Morgan Chase Kaggle Kickstarter las vegas airport lasso regression Lead Data Scienctist Lead Data Scientist leaflet league linear regression Logistic Regression machine learning Maps market matplotlib Medical Research Meet the team meetup methal health miami beach movie music Napoli NBA netflix Networking neural network Neural networks New Courses NHL nlp NYC NYC Data Science nyc data science academy NYC Open Data nyc property NYCDSA NYCDSA Alumni Online Online Bootcamp Online Training Open Data painter pandas Part-time performance phoenix pollutants Portfolio Development precision measurement prediction Prework Programming public safety PwC python Python Data Analysis python machine learning python scrapy python web scraping python webscraping Python Workshop R R Data Analysis R language R Programming R Shiny r studio R Visualization R Workshop R-bloggers random forest Ranking recommendation recommendation system regression Remote remote data science bootcamp Scrapy scrapy visualization seaborn seafood type Selenium sentiment analysis sentiment classification Shiny Shiny Dashboard Spark Special Special Summer Sports statistics streaming Student Interview Student Showcase SVM Switchup Tableau teachers team team performance TensorFlow Testimonial tf-idf Top Data Science Bootcamp Top manufacturing companies Transfers tweets twitter videos visualization wallstreet wallstreetbets web scraping Weekend Course What to expect whiskey whiskeyadvocate wildfire word cloud word2vec XGBoost yelp youtube trending ZORI