Agile Computational Intelligence for Supporting Hospital Logistics During the COVID-19 Crisis

Abstract

This chapter describes a case study regarding the use of ‘agile’ computational intelligence for supporting logistics in Barcelona’s hospitals during the COVID-19 crisis in 2020. Due to the lack of sanitary protection equipment, hundreds of volunteers, the so-called “Coronavirus Makers” community, used their home 3D printers to produce sanitary components, such as face covers and masks, which protect doctors, nurses, patients, and other civil servants from the virus. However, an important challenge arose: how to organize the daily collection of these items from individual homes, so they could be transported to the assembling centers and, later, distributed to the different hospitals in the area. For over one month, we have designed daily routing plans to pick up the maximum number of items in a limited time—thus reducing the drivers’ exposure to the virus. Since the problem characteristics were different each day, a series of computational intelligence algorithms was employed. Most of them included flexible heuristic-based approaches and biased-randomized algorithms, which were capable of generating, in a few minutes, feasible and high-quality solutions to quite complex and realistic optimization problems. This chapter describes the process of adapting several of our ‘heavy’ route-optimization algorithms from the scientific literature into ‘agile’ ones, which were able to cope with the dynamic daily conditions of real-life routing problems. Moreover, it also discusses some of the computational aspects of the employed algorithms along with several computational experiments and presents a series of best practices that we were able to learn during this intensive experience. © 2021, The Author(s), under exclusive license to Springer Nature Switzerland AG.