Integrating human infrastructure in post-disaster critical lifeline restoration scheduling

Abstract

Critical societal services, such as health care, are offered from facilities that are supported by civil lifelines. In disaster events, restoration of the damaged lifeline elements and facilities is necessary for restoring societal function. It is generally presumed that the critical societal services are intact as long as the buildings from which they are provided are functioning and have operating supporting lifelines. This paper recognizes that the services require not only a functioning space for supporting related operations, but also depend on the presence of skilled personnel. For this purpose, a mathematical technique for modeling this human infrastructure as a civil lifeline is proposed. This human infrastructure lifeline modeling approach is integrated into a stochastic program that seeks an optimal schedule of post-event restoration actions across lifelines to minimize the amount of time that civil lifeline elements, including personnel, and the critical societal services they support are nonfunctioning. The model is cast in a real-time setting, where situational awareness increases over time as information is obtained either endogenously via the repair crews or by way of exogenously provided information from other sources (e.g., inspectors, emergency responders, drones, remote sensing). The approach is also useful in estimating critical societal systems resilience. Numerical experiments were conducted on a synthetic, yet realistic problem instance to illustrate the capabilities of the proposed techniques and investigate the importance of adequately capturing the multi-faceted role of personnel in post-disaster service capacity restoration.