Development of Intelligent Systems in 5.0 Supply Chains using Unmanned Aerial Vehicles (#928)

Authors

Elorza López, Neftalí

Vázquez Alonso, Carlos Alberto

González Rodríguez, Carolina

González Villela, Víctor Javier

Macedo Chagolla, Fernando

Sánchez Ruiz, Luis Armando

Abstract

This project is the result of the current biological crisis where it was required to deliver medical supplies to over 15 hospitals in the metropolitan area of Mexico City. During its development, it had the support of Amazon due to its award-winning participation within the health category in the national event called "UNAM - AWS Hackathon" This allowed us to learn more tools for the optimal implementation of artificial intelligence and machine learning algorithms. The project design was to apply digital transformation to deliver supplies anywhere in the world, no matter if an implication exists, such as a natural disaster or a public health emergency. Our objective is to apply digital solutions by implementing Intelligent Systems in the supply chain to cover the demand for a product, doing the delivery process through pilotless aerial vehicles, optimizing delivery times and reducing transportation costs as well as reducing the environmental impact that the process entails. To achieve this objective, we implement software tools so that the vehicle travels safely, safeguarding the integrity of the cargo. We determined Checkpoints in strategic geographical points, complying with the health security and control needs that arise, checkpoints automatically implemented using K-Nearest Neighbors (machine learning model used since due to its characteristics and behaviour is the learning model that best fits our physical model since the control variables such as distances allow us to find the closest point), creating the algorithm for any city in the world. In a different way. We use Amazon Web Services (AWS) because they are easy-to-implement tools and have enough reach globally; however, it is possible to implement it on any other cloud service platform (such as Google Cloud Platform, Microsoft Azure, etcetera). These services have a stable infrastructure and provide us with all the software tools we need to fulfill our objective. The results from the intelligent system showed us that its implementation within pilotless aerial vehicles (quadcopter drone) is viable and managing to satisfy all the needs raised at the beginning of the project due to financial limitations must run out through a simulator.