The aim of this project is to lay down the foundations of a novel approach to real-time control of networked cyber-physical systems (CPS) that leverages their cooperative nature. Most networked controllers are not implementable over embedded digital computer systems because they rely on continuous time or synchronous executions that are costly to enforce. These assumptions are unrealistic when faced with the cyber-physical world, where the interaction between computational and physical components is multiplex, information acquisition is subject to error and delay, and agent schedules are asynchronous. Even without implementation obstacles, the periodic availability of information leads to a wasteful use of resources. Tuning controller execution to the task at hand offers the potential for great savings in communication, sensing, and actuation. The goal of this project is to bring this opportunity to fruition by combining event- and self-triggered control ideas into a unified approach that inherits the best of both models. The key conceptual novelty is for agents to make promises to one another about their future states and warn each other if they later decide to break them. The information provided by promises allows agents to autonomously determine when fresh information is needed, resulting in an efficient network performance. Networked cyber-physical systems are transforming the way society interacts with the physical world. Advances in this field are extending the range of human capabilities in an increasing number of areas with high societal and economic impact, such as smart energy, intelligent transportation, advanced manufacturing, health technology, and the environment. This project contributes to the science and technology of cyber-physical systems by developing a novel principled approach for networked systems to operate efficiently and cope with the sources of uncertainty present in real-word applications. The potential benefits are real-time operation in a wide range of application domains of cooperative cyber-physical systems with a superior level of efficiency and robustness than currently possible. The project promises to contribute to the training of a new generation of engineering students at UC San Diego with the skills necessary to deal with this type of multi-faceted systems and applications. The plan includes undergraduate student involvement in research, graduate supervision and curriculum development, outreach to high-school students, retention of minorities in STEM disciplines, and broad dissemination activities.

Large-scale critical infrastructure systems, including energy and transportation networks, comprise millions of individual elements (human, software and hardware) whose actions may be inconsequential in isolation but profoundly important in aggregate. The focus of this project is on the coordination of these elements via ubiquitous sensing, communications, computation, and control, with an emphasis on the electric grid. The project integrates ideas from economics and behavioral science into frameworks grounded in control theory and power systems. Our central construct is that of a ?resource cluster,? a collection of distributed resources (ex: solar PV, storage, deferrable loads) that can be coordinated to efficiently and reliably offer services (ex: power delivery) in the face of uncertainty (ex: PV output, consumer behavior). Three topic areas form the core of the project: (a) the theoretical foundations for the ?cluster manager? concept and complementary tools to characterize the capabilities of a resource cluster; (b) centralized resource coordination strategies that span multiple time scales via innovations in stochastic optimal control theory; and (c) decentralized coordination strategies based on cluster manager incentives and built upon foundations of non-cooperative dynamic game theory. These innovations will improve the operation of infrastructure systems via a cyber-physical-social approach to the problem of resource allocation in complex infrastructures. By transforming the role of humans from passive resource recipients to active participants in the electric power system, the project will facilitate energy security for the nation, and climate change mitigation. The project will also engage K-12 students through lab-visits and lectures; address the undergraduate demand for power systems training through curricular innovations at the intersection of cyber-systems engineering and physical power systems; and equip graduate students with the multi-disciplinary training in power systems, communications, control, optimization and economics to become leaders in innovation.

Multicore platforms have the potential of revolutionizing the capabilities of embedded cyber-physical systems. Unfortunately, when such systems have safety-critical components, multicore platforms are rarely used. The reason is a lack of predictability associated with hardware components such as caches, memory controllers, etc., that are shared among cores. With current technology, very conservative estimates concerning the usage of these shared resources must be made, to certify that overuse violations do not occur at runtime. The resulting over-provisioning can be significant, easily negating the processing power of any additional cores. The goal of this project is to resolve this multicore "predictability problem" by developing allocation mechanisms that enable shared hardware resources to be controlled in a predictable way. The research agenda in this project includes fundamental research on relevant real-time resource allocation problems, prototyping efforts involving real-time operating systems and middleware, and experimental evaluations of improvements enabled by the developed mechanisms in timing analysis tools (which are used to determine task execution-time budgets). Addressing the "predictability problem" associated with multicore platforms would be a breakthrough result for safety-critical, cyber-physical systems in domains such as avionics and automobiles. When using multicore platforms to host highly-critical workloads in these domains, the current state of the art is to obviate the predictability problem by turning off all but one core. Unless a more intelligent solution can be found, such domains will not benefit from savings in size, weight, and power (SWaP) and gains in functionality that multicore platforms afford. Broader impacts include joint research with industry colleagues on supporting real-time workloads in unmanned air vehicles, the development of publicly-available open-source software that can be used by other institutions for research and teaching purposes, and the development of a new course on cyber-physical systems.

Large-scale critical infrastructure systems, including energy and transportation networks, comprise millions of individual elements (human, software and hardware) whose actions may be inconsequential in isolation but profoundly important in aggregate. The focus of this project is on the coordination of these elements via ubiquitous sensing, communications, computation, and control, with an emphasis on the electric grid. The project integrates ideas from economics and behavioral science into frameworks grounded in control theory and power systems. Our central construct is that of a ?resource cluster,? a collection of distributed resources (ex: solar PV, storage, deferrable loads) that can be coordinated to efficiently and reliably offer services (ex: power delivery) in the face of uncertainty (ex: PV output, consumer behavior). Three topic areas form the core of the project: (a) the theoretical foundations for the ?cluster manager? concept and complementary tools to characterize the capabilities of a resource cluster; (b) centralized resource coordination strategies that span multiple time scales via innovations in stochastic optimal control theory; and (c) decentralized coordination strategies based on cluster manager incentives and built upon foundations of non-cooperative dynamic game theory. These innovations will improve the operation of infrastructure systems via a cyber-physical-social approach to the problem of resource allocation in complex infrastructures. By transforming the role of humans from passive resource recipients to active participants in the electric power system, the project will facilitate energy security for the nation, and climate change mitigation. The project will also engage K-12 students through lab-visits and lectures; address the undergraduate demand for power systems training through curricular innovations at the intersection of cyber-systems engineering and physical power systems; and equip graduate students with the multi-disciplinary training in power systems, communications, control, optimization and economics to become leaders in innovation.

Multicore platforms have the potential of revolutionizing the capabilities of embedded cyber-physical systems. Unfortunately, when such systems have safety-critical components, multicore platforms are rarely used. The reason is a lack of predictability associated with hardware components such as caches, memory controllers, etc., that are shared among cores. With current technology, very conservative estimates concerning the usage of these shared resources must be made, to certify that overuse violations do not occur at runtime. The resulting over-provisioning can be significant, easily negating the processing power of any additional cores. The goal of this project is to resolve this multicore "predictability problem" by developing allocation mechanisms that enable shared hardware resources to be controlled in a predictable way. The research agenda in this project includes fundamental research on relevant real-time resource allocation problems, prototyping efforts involving real-time operating systems and middleware, and experimental evaluations of improvements enabled by the developed mechanisms in timing analysis tools (which are used to determine task execution-time budgets). Addressing the "predictability problem" associated with multicore platforms would be a breakthrough result for safety-critical, cyber-physical systems in domains such as avionics and automobiles. When using multicore platforms to host highly-critical workloads in these domains, the current state of the art is to obviate the predictability problem by turning off all but one core. Unless a more intelligent solution can be found, such domains will not benefit from savings in size, weight, and power (SWaP) and gains in functionality that multicore platforms afford. Broader impacts include joint research with industry colleagues on supporting real-time workloads in unmanned air vehicles, the development of publicly-available open-source software that can be used by other institutions for research and teaching purposes, and the development of a new course on cyber-physical systems.

Continuous real-time tracking of the eye and field-of-view of an individual is profoundly important to understanding how humans perceive and interact with the physical world. This work advances both the technology and engineering of cyber-physical systems by designing an innovative paradigm involving next-generation computational eyeglasses that interact with a user's mobile phone to provide the capability for real-time visual context sensing and inference. This research integrates novel research into low-power embedded systems, image representation, image processing and machine learning, and mobile sensing and inference, to advance the state-of-art in continuous sensing for CPS applications. The activity addresses several fundamental research challenges including: 1) design of novel, highly integrated, computational eyeglasses for tracking eye movements, the visual field of a user, and head movement patterns, all in real-time; 2) a unified compressive signal processing framework that optimizes sensing and estimation, while enabling re-targeting of the device to perform a broad range of tasks depending on the needs of an application; 3) design of a novel real-time visual context sensing system that extracts high-level contexts of interest from compressed data representations; and 4) a layer of intelligence that combines contexts extracted from the computational eyeglass together with contexts obtained from the mobile phone to improve energy-efficiency and sensing accuracy. This technology can revolutionize a range of disciplines including transportation, healthcare, behavioral science and market research. Continuous monitoring of the eye and field-of-view of an individual can enable detection of hazardous behaviors such as drowsiness while driving, mental health issues such as schizophrenia, addictive behavior and substance abuse, neurological disease progression, head injuries, and others. The research provides the foundations for such applications through the design of a prototype platform together with real-time sensor processing algorithms, and making these systems available through open source venues for broader use. Outreach for this project includes demonstrations of the device at science fairs for high-school students, and integration of the platform into undergraduate and graduate courses.

The objective of this research is to address issues related to the platform revolution leading to a third generation of networked control systems. The approach is to address four fundamental issues: (i) How to provide delay guarantees over communication networks to support networked control? (ii) How to synchronize clocks over networks so as to enable consistent and timely control actions? (iii) What is an appropriate architecture to support mechanisms for reliable yet flexible control system design? (iv) How to provide cross-domains proofs of proper performance in both cyber and physical domains? Intellectual Merit: Currently neither theory nor networking protocols provide solutions for communication with delay constraints. Coordination by time is fundamental to the next generation of event-cum-time-driven systems that cyber-physical systems constitute. Managing delays and timing in architecture is fundamental for cyberphysical systems. Broader Impact: Process, aerospace, and automotive industries rely critically on feedback control loops. Any platform revolution will have major consequences. Enabling control over networks will give rise to new large scale applications, e.g., the grand challenge of developing zero-fatality highway systems, by networking cars traveling on a highway. This research will train graduate students on this new technology of networked control. The Convergence Lab (i) has employed minority undergraduate students, including a Ron McNair Scholar, as well as other undergraduate and high school researchers, (ii) hosts hundreds of high/middle/elementary school students annually in Engineering Open House. The research results will be presented at conferences and published in open literature.