Air transport industry has greatly benefited from the increase in travel demand over the last few decades. While airlines and airports’ aeronautical functions are continuously under the limelight with the ever- increasing travel demand, this paper focuses on the airports’ retail revenue, a crucial part of the non- aeronautical functions of the airport which is often neglected despite its importance. According to Airport Council International (ACI 2017), airport retail (including food & beverage (F&B)) is contributing to 32.6% of the total non-aeronautical revenue and it increased by 11% compared to the result in 2015. Airport generated an estimated $19.5 Billion on retail alone in 2016 based on the ACI report (ACI 2016). In fact, some even label airport retail as the Formula 1 of retail because airport generates 36% of its revenue from 16% of the space used on retail.
There is some degree of similarity between airport retail and retail in a shopping mall. However, airport retail has far more restrictions than retail in a shopping mall. The primary objective of passengers is flying, and shopping is not the primary goal in the airport context. As a result, airport terminals are usually designed for its aeronautical functions, even though the design of airport terminal has a significant impact on airport retail which contributes to almost 14% of the overall airport revenue.
Terminal construction is capital intensive, and any mistake will be costly to recover. One feasible way to test out the terminal design without a high cost is through the means of simulation. Current airport simulation models in the literature tend to be discrete-event driven and focus on the process-driven activities such as check-in. Queueing and routine processes are critical elements in discrete event simulation. Meanwhile, airport retail usually does not involve fixed procedures and passengers have their own threads of control. Thus, Agent-Based Model (ABM) could be a more suitable alternative. Agents in the model should be given the full autonomy in making their own micro decisions. However, such an ABM is currently rare among existing airport simulation studies.
This paper aims to demonstrate the potential of ABM in investigating the impact of terminal design on retail revenue through the simulation of passenger’s shopping behavior in different scenarios of terminal design. Such an ABM can contribute to the simulation research in several ways. It is a novel approach to test the relationship between airport retail performance against terminal designs. More importantly, the model demonstrates that realistic agents, who behave like real passengers and interact with the airport retail environment, is possible. The model also provides additional insights into how people could be modeled under an enclosed environment. The implementation of such ABM suggests its potential that can be applied in a broader context, such as shopping mall, stadium or museum. This model serves as a cornerstone for future model expansions.