This is a project presented by Revathy Ramanan, Asmita Mali, Sridhar Ramanathan, and Sudharshan S Iyer, PGP DSBA students, in the AICTE Sponsored Online International Conference on Data science, Machine learning and its applications (ICDML-2020). A follow-up paper was published in the conference journal.
Aviation oil companies supply on-demand fuel to flights; technically referred to as Into-Plane Fuelling (ITP). In the absence of any advance information of fuel uplift volume, a certain aviation oil company employs heuristic approaches to manage the operations leading to considerable variation in the average fuelling time clocked across engineers, further resulting in underutilization of resources and overtime compensation. There have been quite a few researches on optimizing crew assignment but most of which have been from the perspective of the airlines and not from the viewpoint of the aviation oil companies. The objective of this study was to optimize the work efficiency and shift operations of an Aviation Fuelling Company in India by analysing the past fuelling records of the company.
The study was done on data provided by the aviation oil company for one of the busiest airports in India. The sample data consisted one-year airline operational data containing 63000 airplane fuelling records with information collected on 37 variables. In the project, a 6-months flight schedule was also used to stimulate work allocation.
The study proposed an optimized work allocation schedule by spreading out fuelling hours equally across the engineers. The study further proposed a mixed-integer nonlinear optimisation model framework for allocating engineers for fuelling aircraft, considering the constraint particular to this activity where the allocation of first flight to an engineer would be done on First Come First Serve basis (FCFS). The fuel time per engineer ranged from 50 to 250 minutes per day. Such imbalance in time distribution was observed throughout the study period. This model resulted in a potential reduction of variance in fuelling time distribution across engineers by 70%. This will lower the inconsistency of work distribution which will reduce overtime compensation given to the fuelling engineers. The proposed optimisation schedule can be used as a starting point for shift managers and can be tweaked to adjust for any unresolvable clashing flight scenarios, non-scheduled flights, flight delays, or any other operational issue. The model can find a wide application across airport terminals in India and even across the globe.
Wish to work on such interesting capstone projects and learn new concepts in Data Science? Upskill with Great Learning’s PGP Data Science and Business Analytics Course today and power ahead in your career. Feel free to leave your comments below in case of any queries.
