This study takes a unique approach to investigating real-world automated aircraft activity data. The research methodology was implemented, with data collected via wireless data transfer from an Airline Information Management System (AIMS) archive and manually collected Passenger Statistics (PAX)
data from aircraft operated by Aero Contractors Company (Lagos, Nigeria). This data has not been fully explored in previous research. Addressing the existing gaps in this area aims to provide a deeper understanding of the potential trade-offs and challenges associated with using such automated onboard data within an airline.
The initial findings revealed that the AIMS data did not contain associated or allied data, such as passenger statistics; this necessitated the merging of such passenger data with the automated data. The process of merging the AIMS and PAX datasets using take-off timestamps achieved an impressive 60% merging efficiency, which, while reasonable, excluded a significant proportion of atypical
flights, most notably those associated with delays. An analysis of the alignment performance indicated that the failure to align more AIMS and PAX correctly was due to the two datasets logging different take-off time indicators. A strategy was adopted, allowing a few AIMS-PAX take-offs of up to 1 hour, improving the merging efficiency to over 85%. Furthermore, increases in time with a higher cut-point achieved a merging efficiency better than 90%. However, increasing this difference also increases the potential for misalignment and duplicate alignments. A rolling fit strategy was devised that first fit the
nearest AIMS and PAX flights. Then, excluded these before extending the allowed time difference forward and backwards and locating the next most likely fits, and the results achieved a better than t92% merging efficiency with negligible multiple matches of 3 out of 46,912 or less than 0.0001% of
cases in the resulting merged dataset. This better merging efficiency across the ranges of other parameters in both datasets arguably demonstrates the more robust nature of the merging strategy.
Early investigations provide insight into model-derived estimates of the fuel for ‘Single Aisle Jet’ types, known as narrow-body commercial airliners, was carried out. The meta-analysis suggests that
flight distance ranges, aircraft types, and modelling techniques all play a significant role, to varying
degrees, in explaining the variations in parameter estimates, model-derived fuel estimates (fuel burn) and flight distance. They suggest that the model-derived fuel estimates per flight distance travelled for narrow-body commercial airliners operating short and medium-haul flight aircraft types decrease with increasing flight distance. This trend is observed across different flight distance ranges, with flights over 500 km suggesting lower model-derived fuel estimates. However, it is important to note that a non-linear association between model-derived fuel estimates and flight distance was found, indicating a more complex relationship. Furthermore, the findings of this research demonstrate a positive
association between model-derived fuel estimates and flight distance travelled for Boeing 737s.
Accordingly, models such as the International Civil Aviation Organisation (ICAO) Fuel Estimation Tool (IFSET) were investigated. The IFSET model may be used with flight distance and flight time as input. The performance of IFSET was investigated, and results show small overestimates of about 4%
and 17% for comparisons of AIMS measurements with distance and time-based IFSET predictions, respectively. To address these issues, the IFSET model was recalibrated. The recalibration process involved adjusting the model’s parameters to fit the actual data better, thereby reducing the bias that
distorts the comparison of fleets operating over different distances and times. Recalibration resulted in a significant reduction in IFSET bias, demonstrating the importance of this continuous improvement process. In addition, aircraft-related effects were quantified, suggesting some, most likely small,
aircraft-to-aircraft variation in fuel economy. The majority of the B737-500s (i.e., 5N-BKQ, 5N-BKR, 5N-BLC, 5N-BLD, 5N-BLE, and 5N-BLG) with similar engine types (CFM International-CFM56-3C1) show slightly lower fuel consumption rates (4.01–4.19 kg/km). In contrast, B737-400s (5N-BIZ and 5N-BJA) with similar engine types (CFM International-CFM56-3B2) – were moderately at 4.48
kg/km and 4.56 kg/km, respectively, and one, a B737-400 (5N-BOC) with a similar engine to the B737-500s (CFM International-CFM56-3C1), showing 4.87 kg/km.
Additionally, aircraft engine thrust ratings were extracted from the CFM56-3 Engine Handbook and merged with the AIMS automated data for each aircraft. The OAT from the National Oceanic and Atmospheric Administration (NOAA) database was then integrated with the AIMS dataset based onflight date (AIMS) / date (OAT) and taxi-out time (AIMS) / hourly window (OAT). Notably, outside
air temperature (OAT) and EGT emerged as pivotal factors in model performance, overshadowing flight time. This insight, gleaned from the comparison of spline term response ranges, is a significant contribution to understanding EGT as it affects aircraft fuel consumption, providing a deeper
understanding of the factors influencing aircraft fuel consumption. EGT predictions suggest a near-linear EGT gradient of 3 °C for a 1 °C rise in OAT for the fleet under investigation when compared with a gradient of 3.2 °C for every 1 °C of OAT reported in the CFM56-3 Aircraft Owners and Operators Handbook. Also reported in the CFM56-3 Aircraft Owners and Operators Handbook is an EGT of 32 °C at an OAT of 20 °C compared to an OAT of 10 °C.
However, EGT predictions are 40 °C at an OAT of 20 °C compared to an OAT of 10 °C. The resulting steep gradients underscore the pressing need for a deeper understanding of the EGT and fuel consumption relationship in aircraft operating in warmer climates, such as the Aero Contractor fleet.
This study’s findings suggest that using automatic onboard logging data can be beneficial. Automatic onboard logged data were found to aid in positively understanding the sources of fuel economy. The implications of these findings are significant for the aviation industry, as they can inform strategies for improving fuel efficiency and reducing environmental impact. Specifically, the research suggests that the industry should consider incorporating automated data into their fuel efficiency strategies, as this could lead to significant cost savings and environmental benefits.