Collective intelligence operations of human and virtual agents in CPS (Cyber Physical System)

Abstract

This thesis deals with contemporary emergent approaches to CPS (Cyber Physical System) cognitive automation and embedded-intelligence processes, either with or without a HitL (Human-in-the-Loop) setting. The introduction section describes the concepts of normal processes vs. process exceptions, with their related research questions (read section 1.1 at page 2). Chapter two deals with the most relevant issues that affect CPS control systems, with general descriptions of their technical aspects. Chapter three is centered around cognitive technologies implemented in CPPs (Cyber-Physical Processes) automation and control, with some specific application examples. Chapter four deals with what kind of AI (Artificial Intelligence) aspects need to be tackled so to progressively improve intelligent automation and control systems to a human-like cognition level. The final chapter relates to synthesizing a list of best practices for the discussed AI-driven technologies, in order to design, implement, test, manage and monitor a reliable CPS system, according to its main application domain.