This is a converted executive summary from the Bachelor Thesis of yaos data analyst, Michel Baud. The thesis analysed historical data and airline specific input factors in order to improve load planning.

Our data analyst, Michel examined the potential usage of loading data from yaos web-based application, LimeFlight, using data mining methods. Using historical and freshly generated data, yaos aims to optimise it’s web application LimeFlight to offer airlines the ability to reduce and optimise their loading plans. They intend to do this by adding a function for recommending load quantities of a flight segment. Such a function is unprecedented in the airline catering industry and could significantly improve the way that airlines plan and execute their loadings in the future.

Michel used the CRISP-DM method to carry out his research. This involves using data mining methods implemented with predefined phases and partial tasks. The results clearly revealed that load patterns can be seen in the data, and that loadings differ from flight-specific data such as flight time, seat classes, number of passengers and the type of aircraft. This knowledge makes it possible to describe a load by the flight-specific data and thus to calculate loading quantities.

In the study, Michel included four sample groups: two short-distance, one middle-distance and one long-distance loading type. Michel discovered that such a sampling is insufficient to calculate a solid mathematical formula for loading quantities. In the future further subgroups of each sample would need to be identified to address individual factors such as different departure times and the associated meals. Further descriptors are also necessary to conduct a more thorough analysis.

Michel’s research revealed that there was a clear opportunity for airlines to leverage existing data to optimise their processes. To find and decipher the further descriptors and subgroups required, Michel joined the yaos team as a data analyst in August 2016. His role is to delve deeper into the possibilities and add a new, innovative and unprecedented dimension to the LimeFlight application. Watch this space.

Want to join the discussion? Leave your comments below and spread this post onLinkedIn, Twitter and Facebook using our handle @yaos_ch.