MAS4CHICKEN: Multi-Agent Systems based order service model for waiting time reduction at rotisserie chicken restaurants in Lima - Peru

Melendez N.; Cruz J.; Salas C. · Proceedings of the International Conference on Soft Computing and Machine Intelligence, ISCMI (2024) · DOI: 10.1109/iscmi63661.2024.10851627

Resumen

A very frequent problem at rotisserie chicken restaurants in Lima - Peru is the diners' long waiting times. This generates discomfort on the diners, as well as negative reviews towards these kinds of restaurants. There are studies out there that aim to combat this issue by using simulating software such as Arena, Promodel, AnyLogic, etc. However, this article uses a different method called 'Multi-Agent Systems (MAS)'. In order to combat the aforementioned issue, the MAS4CHICKEN model has been created, which aims to determine the optimal scenario in which the diners are attended with the lowest delay possible, allowing for the diners to feel satisfied. This implemented method mainly uses three different kinds of agents: Waiter, Chef, and Diner. This model was developed with the NetLogo modeling software, which makes it easier to work on Multi-Agent Systems and their characteristic behaviors. The proposed model was tested under different scenarios. By the end of the tests it was concluded that 4 chefs are able to optimally attend diners that come by every 5 minutes, resulting in an attention time of 13.76 minutes. Additionally, it was determined that the model allows the user to evaluate strategies and measure the impact of its implementation in a virtual environment. While this model was firstly designed for rotisserie chicken restaurants, it can be used, either as-is or extended, to work with different kinds of restaurants.

Instituciones: Universidad Peruana de Ciencias Aplicadas