One challenge, that information technology faces, is how to cope with continuous contextual changes and the need for adjustments of the underlying networked system during runtime. In application domains like the Internet of Things (IoT) among others, system adjustments can’t be performed in a device-by-device manner due to a large number of spatially distributed autonomous nodes. As a consequence, innovative approaches for autonomic management are required for such systems. A promising solution for network infrastructures are intent-based approaches. With the idea of declaratively specified goals for network functions, operators can holistically manage the network infrastructure at a high level of abstraction. Anyway, concrete design and implementation concepts are required to enforce intents within a networked system of autonomous nodes in a coordinated manner to ensure consistent behavior of the system according to the declared intent. In this chapter, concepts and approaches of self-adaptive systems are explored as a promising solution to serve as a basis for designing and implementing intent-based networked systems. The authors focus particularly on role concepts that allow a continuous design and implementation of system variability at run time. An example scenario from the IoT domain is used to continuously illustrate concepts and to demonstrate how networked self-adaptive systems can benefit from the introduced role-based concepts.