In automated planning, faceted reasoning is an approach for interactively navigating the solution space of a planning task. This thesis explores the use of faceted reasoning in two settings: explaining learned symbolic action policies and supporting robust observation planning. The first use case addresses the explanation and analysis of learned symbolic action policies, a learning-based approach to solving classical planning problems. While policies are in principle human readable, understanding how exactly they solve all planning tasks of a domain, or more importantly, explaining the reasons why they fail is challenging as the logic expressions are often complex. We introduce a framework that allows us to analyze policies learned over description-logic features. Each policy rule consists of a set of conditions that have to hold in a state and a set of effects that must occur along the respective transition for the rule to fire. We encode the policy rules in answer-set programming (ASP), extending an established ASP formalism for classical planning. Using faceted reasoning, we analyze how individual policy rules affect the space of possible plans and identify issues in the policy when it fails to solve a task. Our analysis is feasible on tasks that can be solved optimally by strong classical planners. The second use case addresses robust observation planning. As an example, we consider the task of scheduling the operations of a satellite constellation to take pictures of a set of regions of interest and send these observations down to Earth. With large constellations, uncertain observation conditions, e.g., due to clouds and limited availability of downlink capacity, it becomes increasingly hard to coordinate the operations. We address the setting where satellites have limited memory and pictures need to be sent through the constellation to a satellite that can transmit them to a ground station. A key challenge in this problem is the highly dynamic environment, which requires a planning system that can quickly adapt to changes. Here, faceted reasoning allows the current solution to be dynamically updated as conditions change and avoids replanning from scratch. We show empirically that facets offer a scalable approach for constellation planning and illustrate the robustness in a case study.