Smart Cyber-Physical Systems

Smart Cyber-Physical Systems and Internet of Things are some of the names attributed to the next generation of systems consisting of a large number of distributed and potentially mobile components that interact with their environment and with other components of the system. Thanks to this extensive collaboration and data sharing, the whole system becomes “smart” in its decisions. This means that the system can adapt to various situations and derive the best out of them. An important feature of these systems is that their “smartness” increasingly relies on software, up to the point when the software becomes by far their most complex and most critical part. Another important feature of these systems is that they are highly dynamic – as components of the system move and adapt to new situations in their environment, the whole system has to continuously reconfigure itself.

The resulting software complexity is typically tackled by building hierarchical software architecture abstractions. However, traditional methods of modeling software architectures cannot easily cope with the level of dynamicity present in these systems. In this lecture, we will look into novel software architecture concepts that specifically target the dynamicity and allow tackling the software complexity of the systems. In particular, we will introduce a concept of autonomous component ensembles, which provide a hierarchical abstraction for modeling dynamic coalitions of components.