In this talk, I’ll describe a succession of complex systems modelling within different scientific disciplines, such as financial and banking networks, immunology, chemistry, climatology, with, in all cases, the pleading for object orientation and the resorting to UML diagrams to facilitate the interfacing between the software developers and the researchers in those disciplines. In a second time, I’ll discuss the concept of emergence, its counter-scientific nature, arguing for a better investigation of the necessary ingredients contributing to a necessary de-subjectivation. For instance, in biology, three of those ingredients could allow a collective phenomenon to be labelled as “emergent.” First, the phenomenon, as usual, requires a group of natural objects entering in a non-linear relationship and potentially entailing the existence of various semantic descriptions depending on the human scale of observation. Second, this phenomenon has to be observed by a mechanical observer instead of a human one, which has the natural capacity for temporal or spatial integration, or both. Finally, for this natural observer to detect and select the collective phenomenon, it needs to do so on account of the adaptive advantage this phenomenon is responsible for.
Self-adaptation equips a computing system with a feedback loop. This feedback loop resolves uncertainties that were difficult to anticipate before the system was deployed, such as sudden changes in the availability of resources and fluctuating workloads. One of the key challenges of engineering self-adaptive systems is ensuring that the system complies with the adaptation goals, regardless of the uncertainties. A prominent approach to tackle this challenge is the use of formal approaches at runtime. In this talk we explain the basic principles of self-adaptation and zoom in on decision-making using formal approaches. We illustrate these principles for ActivFORMS (Active FORmal Models for Self-adaptation), an approach that we developed in our research group. ActivFORMS spans four main stages of the life cycle of a feedback loop: design, deployment, runtime adaptation, and evolution. We give an overview of ActivFORMS and zoom in on one of its distinct features: the use of statistical model checking of runtime models to select adaptation options that realise the adaptation goals with a required level of accuracy and confidence. We use an Internet of Things application for building security monitoring as illustrative case.