Design of Experiment (DOE) Analysis on Aviation Fuel Efficiency

Problem Statement

Airlines incur significant operating costs due to fuel consumption, and fuel efficiency plays a major role in their profitability. Multiple factors, such as aircraft type, altitude, weather conditions, and route distance, all contribute to fuel usage.

For my capstone project in IE533, I conducted an investigation using Design of Experiments (DOE) to understand how the type of aircraft and altitude affect the fuel consumption of an airline operating across three different routes. The goal was to optimize airline operations, reduce operating costs, and improve overall efficiency by analyzing the influence of aircraft type and altitude.

Key Factors Studied:

Aircraft Type: Two different aircraft models
Route: Two specific routes
Altitude: Two altitude levels

Fig 1: Power Analysis: Determines the sample size as 64

Experiment Design

For this project, I designed the experiment as a 2³ factorial randomized complete block design. The three factors studied—aircraft type, route, and altitude—were randomized within each block, with eight total flights (one for each combination of the factors). To eliminate confounding factors, the order of flights was randomized, ensuring no aircraft type was flown twice in a row on the same route and altitude combination. Replication was used to improve precision, and a power analysis, conducted using Minitab, helped determine the required sample size [64]. To ensure accurate results, variables such as weather, cargo, and flying speed were controlled.

Data Setup

For this experiment, data was collected from flight factors, including altimeter readings from the airplanes, pilot inputs for the route, and aircraft flight logs. A total of 64 flight records were gathered, with 8 replicates for each aircraft type, route, and altitude combination. To ensure that no confounding variables were introduced, no flights involving the same aircraft type, route, and altitude were conducted on the same day.

The response variable, fuel consumption, was generated using an equation that applied appropriate weights to each factor:

Response Equation: 50000 - ('Altitude' * 0.25) + IF('Aircraft Type' = 2, -50, 10) + IF('Route' = 1, 7000, 7500) + 10 * IF('Route' = 2, 1000, 500) * IF('Aircraft Type' = 2, 1, 0.5) + 'Noise'

Noise was incorporated into the model to account for inherent variability in the data, represented by a triangular distribution:

Noise Equation: Noise ~ Triangular(-500, 0, 500)

Analysis

The null hypothesis for the experiment would be that three factors have a significant effect on the fuel consumption of an aircraft.

Ho: 𝜏 (Aircraft Type) = 𝜏 (Altitude) = 𝜏 (Route) = 0
Ha: at least one τ is not equal to 0 (α=0.05)

After testing all model assumptions, I ran ANOVA and T Test to conduct hypothesis testing.

Fig 2: Hypothesis Testing using ANOVA and the T Test

I determined through the ANOVA table and accompanying charts that all main effects—Aircraft Type, Route, and Altitude—are statistically significant. Additionally, the interaction between Aircraft Type and Route was found to be significant.

Based on these results, I rejected the null hypothesis and concluded that Aircraft Type, Route, and Altitude play a crucial role in influencing overall fuel consumption, directly impacting the operational costs of the airline. Furthermore, the model proved to be robust, as the assumptions were validated by the assumption plots, ensuring the reliability of the analysis.

Conclusion

Through this experiment, I determined that Aircraft Type and Altitude had a significant impact on fuel consumption across the airline’s two routes. Among these factors, Altitude played the most critical role in reducing fuel consumption and overall operating costs—higher altitudes resulted in lower fuel usage.

Additionally, Aircraft Type 1 was found to be more fuel- and cost-efficient compared to Aircraft Type 2 on both routes. Based on these findings, I recommended that the airline operate Aircraft Type 1 and fly at the highest feasible altitude (35,000 ft) to minimize fuel consumption and reduce operating costs.

Further experiments could be conducted to explore additional factors that might contribute to cost reductions and operational efficiency for the airline.

Page updated

Google Sites

Report abuse