The recent trend of uncontrolled spread of infectious diseases has resulted in severe disruption to society on a worldwide scale. One of the causes is represented by public transportation services that contribute to the spread of an epidemic, which has consequences for human health and the economy. Public health organisations call for support from scientists to identify the main criticalities and take countermeasures on time.

This paper aims to support public authorities by exploiting the capability of self-adaptive systems (SaS) to autonomously modify their behaviour when subject to changes in their environment. We inherit from the literature patterns that integrate distributed and central control of SaS, and we demonstrate the effectiveness of these design patterns when deciding public health measures applied to transportation services during an epidemic.

Our novel methodology consists of modelling and analysing control action types that describe a unique interaction between a central controller and a distributed controlled transportation system to get a desired adaptation. We rely on probabilistic model checking to provide formal guarantees on the expected number of infections determining the epidemic evolution.

Experimental results show that our technique is adequate to counteract the epidemic scenarios, thus supporting public health authorities in monitoring the status of transportation services and making informed decisions.