Engineering collective intelligence at the edge with aggregate processes

Edge computing promotes the execution of complex computational processes without cloud, i.e., on top heterogeneous, articulated, and possibly mobile systems composed IoT edge devices. Such a pervasive smart fabric augments our environment with networking capabilities. This leads to dynamic ecosystem devices that should not only exhibit individual intelligence but also collective intelligence—the ability take group decisions or process knowledge among autonomous units distributed environment. Self-adaptation self-organisation mechanisms are typically required ensure continuous inherent toleration changes various kinds, distribution devices, energy available, load, as well faults. To achieve this behaviour in massively setting like demands, we seek for identifying proper abstractions, engineering tools therefore, smoothly capture behaviour, adaptivity, injection concurrent activities. Accordingly, elaborate notion “aggregate process” computation whose interactions sustained by team spatial region can opportunistically vary over time. We ground extending aggregate model toolchain new constructs instantiate regulate key aspects their lifecycle. By virtue an open-source implementation ScaFi framework, show basic programming examples case studies computing, evaluated simulation realistic settings.