Airlines overbooking

 Introduction

    Companies of all sizes and types all over the world are attempting to optimise profits and

achieve profitable growth in an extremely volatile and competitive business climate. Airlines

are dealing with a challenge that is both economically and environmentally significant. This

problem is handled by implementing an AI system whose goal is to oversell passenger

tickets, as stated below.

The problem

    Dealing with a large number of clients can lead to a variety of issues, but the one we're

interested in is illustrated by the following example: one individual buys a plane ticket, but

for whatever reason, he does not show up at the airport to board the plane. This may not

appear to be a significant issue, but if a plane has a capacity of 100 passengers and 60 of

them fail to show up, the trip will be operated with less than half of its capacity. Because a

consistent portion of customers did not show up, the corporation might possibly generate

more than 60% more profit if they had sold more than 60% more tickets. Additionally, if an

aircraft burns gasoline half-filled, there are environmental issues, since many individuals

may benefit from that specific journey.

The solution

    As previously stated, in that situation, if the airline business had sold 60% more tickets, they

would have maximised their profit, earning an additional 60% profit for one journey.

Obviously, this constant number varies for each flight, and estimating it correctly can

increase the airline's profit. Airlines are able to determine how many seats they should

oversell every trip using a comprehensive artificial intelligence system that takes into

account various aspects such as historical flight data, date, time, even special events, and

much more.

Risky?

    In this paticular situation, from the perspective of an airline company, there are four types of

clients: 

    If the AI system correctly predicts the variables, the A1 and B2 values will be as high as

possible, increasing the company's profit. If the AI fails, however, the A2 and B1 variables

will rise. A greater A2 rating means spoilt seats, while a higher B1 value means the airline

company will lose a lot of money. The B1 value reflects the passengers that the AI algorithm

projected would not show up for the flight, but who actually showed up. With our 100-seat

aeroplane example, this occurs when an airline company sells 110 tickets in the hopes that

only 10 people would show up, but all 110 passengers show up for the flight. This implies

that ten individuals will be unable to fly since there are no available seats. The airline firm will

face a large financial penalty for each and every individual in this case.

Is that greedy?

    Overselling aircraft tickets helps airlines maximise earnings, but it also benefits passengers,

according to industry experts. It provides them more options and access when booking

flights, as well as allowing airlines to charge less for a ticket - which is one of the goals of

airlines, since the more inexpensive tickets are, the more people will buy them.

Workflow

The General workflow can be summarized by the below diagram:

Benefits

The following are the main benefits reaped by airlines that have implemented an Optimal

Overbooking Strategy:

- Reduced spoiled seats, as well as the number of involuntarily denied boardings, which not

only increased revenue for the airline due to additional bookings, but also reduced additional

costs due to compensation due to DNBs (denied boardings) and improved the company's reputation;

- Accurate prediction combined with historical booking patterns assisted the client in

network planning.

Bibliography

https://www.itcinfotech.com/

https://www.youtube.com/channel/UCo7a6riBFJ3tkeHjvkXPn1g