A Study of Architectural Models for Cyber-Physical Systems

点击翻译按钮获取中文摘要

The focus of this project is on cyber-physical systems (CPS) -- physical and engineered systems whose operations are integrated, monitored, and controlled by a computational core. Applications of CPS include, for example, critical (physical) infrastructure control (electric power, water resources, gas and fuel distribution, transportation, etc.), highly dependable medical devices and systems, traffic control and safety, advanced automotive systems, process control, and manufacturing, energy conservation and environmental control, avionics, distributed robotics (tele-presence, tele-medicine), and manufacturing.Traditional frameworks for designing and configuring computing and networking systems do not capture the complexity of CPS systems -- principally because many of the abstractions developed for computing and networking deliberately abstract away essential properties of the physical world. Existing real-time operating systems (mix of operating systems, middleware, and general-purpose virtual machines) are not sufficient to support the complexity challenges of emerging classes of cyber-physical systems because of the mismatch between traditional computation abstractions and properties of physical processes. Some current technical approaches partially bridge the abstraction gap (such as real-time operating systems, middleware technologies, specialized embedded processor architectures, and specialized networks). However, current research and development of cyber-physical systems calls for non-incremental, revolutionary approaches, where new abstractions are built from the ground up. These abstractions must enable reasoning at every level for high-confidence system composition and addresses issues such as time-criticality, safety, and security.In response to the need for a CPS framework, this SGER project will investigate architectural models with the goal of helping the research community to develop a cyber-physical system vision that integrates both computer science (real-time systems; embedded systems; software and system verification, validation, and certification; computing technologies; wireless and wired networking; sensor nets) and the traditional physical sciences and engineering disciplines (control, physical sensors and actuators, physical system design and engineering, materials, even biology and nanotechnology). Such models are likely to provide a platform for future cyber-physical systems design that brings these interacting elements into a single focus.This project will also sponsor several meetings (workshops, symposia, summit, etc) that aim at providing a forum for engineers and scientists in academia, industry and government to present their latest research findings in any aspects of cyber-physical systems, and, consequently, allow the community to define and enhance the research agenda for this important emerging field. The primary intellectual merit of the proposal resides in analyzing existing cyber-physical systems and developing abstracted architecture models that will enable future cyber-physical systems design and realization. The broader impact of the proposed work is on the research community at-large, through the project focus on issues related to cyber-physical systems. The advancement achieved through this Cyber-physical system research is expected to improve the research and development capabilities in the nation and contribute to increased global competitiveness.