Practical, theoretically sound strategies are needed to design recovery strategies that account for heterogeneous users in massively damaged, interdependent infrastructures. This dissertation provides a framework for developing a controller to repair a geographically large interdependent system with many stakeholders that is not suited to straightforward mathematical characterization. The framework employs a satisfactory control approach based on proportional-integral-derivative, or PID, principles, applied to a discrete-time, time-varying system. The proposed framework also combines network science and Systems-Theoretic Process Analysis to design a recovery strategy for interdependent infrastructure. The combination of network science and Systems-Theoretic Process Analysis are complementary, and novel: Systems-Theoretic Process Analysis frames losses in terms of a system, system components, and component interactions [1]. Network science facilitates the depiction of the system, system components, and component interactions using feature-rich edge lists. Additionally, notion of a controlled process in the Systems-Theoretic Process Analysis fits well with the “resilience curve” ([2]) framing of recovery problems. Accordingly, this dissertation exploits the control and feedback framing to shape transient state characteristics to limit impacts to residents and to respond to new information, constraints, and disturbances. This proposed framework was applied to a case study in a real interdependent infrastructure system. Goals were set based on real-world policy considerations, and then formalized into formal control rules using Systems-Theoretic Process Analysis. The PID-based controller that was designed using this methodology produced results that are superior to recovery strategies built without feedback. Graphical techniques to visualize controller outputs and refine them were also provided and demonstrated. This latter set of tools acknowledges the need for stakeholder input, particularly in public sector systems. Finally, an adaptation of the Resilience Matrix approach is provided as a monitoring tool during implementation of the recovery framework. The Resilience Matrix can be used to assemble evidence that recovery is or is not meeting decision-maker goals. The result is a practical and theoretically sound decision support methodology that accounts for key locations, proximities, alternatives, interactions, and sources that constrain system users (the problem structure), as well as individual features and “system properties downscaled” to the component level. This framework is designed to be interoperable with common data formats and asset management systems, and can be integrated into a live work-planning application. This work contributes to efforts to support human population resilience by treating peoples’ resilience as a function of infrastructure system resilience.