This project is aimed at analyzing the factors leading to commercial flight delays, specifically targeting small operating airlines. The goal is to analyze historical flight data from Carrier On-Time Performance Report (2018 — present) on Bureau of Transportation Statistics (BTS)  to draw insights and eventually use this knowledge to develop a machine learning model that predicts the probability of flight delays.
According to Figure 1, Alaska Airlines (AS), Spirit Airlines (NK), and PSA Airlines (OH) have the highest number of delays due to the large volume of flights they operate each month. Despite this, Frontier Airlines (F9), Allegiant Air (G4), Commute Air (C5), and PenAir (KS) experience the highest delay ratios, with PenAir (KS) leading in terms of the highest delay ratio. Therefore, the conclusion for volume and ratio of delays is that smaller airlines (regional/cargo airlines) have higher rate of delayed when comparing with small operating airlines.
Figure 2 reveals a noteworthy pattern: Itinerary with a higher volume of delay incidents tend to exhibit a lower delay rate, and conversely, those with lower delay volumes have a higher rate of delays.
Top 5 routes with highest number of delays are: OGG — HNL (Kahului Airport to Daniel K. Inouye International Airport), SEA — ANC (Seattle-Tacoma International Airport to Ted Stevens Anchorage International Airport), PDX — SEA (Portland International Airport to Seattle-Tacoma International Airport), SEA — PDX (Seattle-Tacoma International Airport to Portland International Airport), HNL — OGG (Daniel K. Inouye International Airport to Kahului Airport).
Top 5 routes with highest ratio of delays: YNG-PIE (Youngstown-Warren Regional Airport to St. Pete-Clearwater International Airport), VPS-SRQ (Destin–Fort Walton Beach Airport to Sarasota-Bradenton International Airport), TRI-AVL (Tri-Cities Airport to Asheville Regional Airport), TPA-MIA (Tampa International Airport to Miami International Airport), STL-BMI (St. Louis Lambert International Airport to Central Illinois Regional Airport at Bloomington-Normal), SFB-YNG (Orlando Sanford International Airport to Youngstown-Warren Regional Airport).
In addition, interstate flights tend to exhibit a higher delay ratio compared to intrastate flights. The majority of delayed flights fall within distance groups 5, 7, 9, and 11. These flights primarily involve interstate routes connecting states within specific regions, based on the distance groupings. However, it’s notable that interstate flights linking states from different regions experience a higher rate of delays. This could be attributed to:
- The intricacies of airspace congestion and routing: longer flights might traverse regions with more complex air traffic control and weather patterns, increasing the likelihood of delays.
- The advantages of shorter distance flights: shorter distance flights may operate within well-established and efficient air corridors, leading to fewer delays.
December, June, July, August, and April stand out as high-demand periods within the aviation industry for several reasons:
- Holiday season (December): December has major holidays such as Christmas and New year’s. Many people travel this time resulting in a surge in air travel demand.
- Summer vacation (June, July, August): These months encompass the peak of summer therefore, the demand for travel increases leading to a significant spike in air traffic.
- Spring Break (April): It is also a good time for travelling in flower season and after winter term, leading to a higher delay ratio.
Monday, Friday and Sunday are days of week with highest delay ratio. The reason for this pattern can be attributed by behaviors of customers in leisure travel in weekend starting from Friday and ending on Sunday. Besides, business trips and workweek travel demand also one of reason that cause the ratio of delays increase on Monday.
The ratio of delays increases in late arrival time (evening and night time) while it increases in noon and early morning for departure time. There are several reasons for this pattern:
- The air traffic volume: During the evening and night air traffic tends to decrease, leading to fewer available flight paths and airspace.
- Weather and Visibility: Weather conditions often deteriorate during the evening and nights, including reduced visibility, fog, and lower cloud ceilings.
- Crew fatigue and scheduling: If a flight is delayed earlier in the day, it can create a ripple effect of crew scheduling challenges that cause delays to accumulate.
- Maintenance and turnaround times: Late evening and nighttime schedules may coincide with maintenance activities and airport operations transitioning.
- Peak Traffic: Noon is a popular time for air travel, often referred to as a “rush hour” for airports.
The majority of delayed flights among small airlines primarily experience delays due to Carrier Delay with NAS (National Aviation System) being the second most common cause (Figure 8). While the available data may not provide extensive insights into the specific reasons for delays, it appears that weather and security factors are not the predominant contributor to these delays. Remarkably, Spirit Airlines (NK) and Alaska Airlines (AS) emerge as the carriers with the highest delayed ratio caused by NAS among all carriers.
Tail number relates to the maintenance, aircraft effectivities and readiness. Figure 8 in Appendix shows top 10 tail numbers of each airline, indicating which aircraft with highest delay ratio. Therefore, airlines can track the maintenance history, and flight records to reduce delay rate when operating aircraft for flight schedule. This is also helpful when airlines consider which aircraft to buy and hire (Table 2).
 “Marketing Carrier On-Time Performance (Beginning January 2018).” Bureau of Transportation Statistics, www.transtats.bts.gov/DL_SelectFields.aspx?gnoyr_VQ=FGK&QO_fu146_anzr=. Accessed 27 Aug. 2023.