Software Technologies for Open, Decentralized and Self-adaptive IoT Systems (PERSIST)

Seizing the immense potential of the Internet-of-Things (IoT) in smart cities and digitized industries, calls for open, decentralized and self-adaptive IoT systems. Yet, current IoT systems are often (1) single-purpose, which hinders the realization of open IoT ecosystems with active participation of all stakeholders, (2) centralized, which limits their applicability in extensive smart city deployments, and (3) require human operators to balance collective concerns, hampering timely and accurate responses. To address these limitations, PERSIST will investigate: “How to (1) specify various stakeholder concerns in an open IoT ecosystem, in particular performance, reliability and safety, (2) translate these concerns to a distributed IoT hardware and software platform, and (3) automatically trade-off these (often diverging) concerns to satisfy the collective behavior of the IoT system?” This project will result in (1) language primitives that allow stakeholders to express their concerns; (2) models and coordination mechanisms to enable systems to autonomously balance stakeholder concerns; and (3) a runtime adaptation approach that combines statistical verification with a set of learning techniques enabling systems to guarantee stakeholder concerns of large-scale IoT ecosystems. The research will be driven and validated in a simulated smart cyber-physical system and a physical IoT testbed infrastructure.

Keywords:Internet-of-Things, Self-adaptation, Coordination

Disciplines:Software engineering, Networking and communications