Workforce and task optimization to guarantee oxygen bottling under a COVID-19 pandemic scenario: A Chilean case study

This work describes an optimization model to analyze the response of a Chilean oxygen-bottling plant to disruptions induced by large external events such as the COVID-19 pandemic. The model addresses employee shift scheduling, machine scheduling with setup times, production scheduling and inventory management, thus being non-linear, although resolved in reasonable computation times. The workers are assumed to have the same training and skills for a specific task, and to behave homogeneously from the psychological and physical perspectives. The objective function minimizes the cost of bottling oxygen, using monthly average of deaths as a proxy variable to model oxygen demand. The results obtained with this model suggest optimizing task assignments as a short-term solution to satisfy an increased oxygen demand representative of a pandemic scenario, while an additional long-term solution is installing a second cryogenic tank. Considering the latter solution, the model allows to quantify the number of workers necessary to operate an enlarged plant. An updated qualifications matrix coupled with an optimized schedule shows that the number of workers required under these conditions does not have to be doubled, but would have to be increased by only one worker.