Experimentation with Self-configuring Systems

The Internet of Things (IoT) has received increased interest over the last years. Objects of the real world (e.g., cars, traffic lights, buildings, mobile phones, etc.) become connected to the Internet. It is expected that self-configuration becomes an increasingly important subject with the IoT. Self-configuring systems (SCS) adapt their behavior to changes in the environment by dynamically change properties, deploying components, or removing them for instance. For this purpose, SCS use policies provided by a requirements engineer. SCS can be found in various application areas. Often, they are used in situations, where maintenance activities cannot be carried out by a human. One reason is lack of time, e.g. the user expects instantaneous adaption to his needs, which is the case with service-oriented systems or some kinds of embedded systems. Runtime representations of requirements (e.g. feature or goal models) are used as policies for SCS. They describe runtime variation points and indicators that can lead to a new configuration. This poster presents a proposal for experimentation with self-configuring systems targeting at both researchers and practitioners in the field. The contribution from th RE perspective is a sandbox to study requirements at runtime.