The proposed decentralized/distributed control and optimization for the critical cyber-physical networked infrastructures (CPNI) will improve the robustness, security and resiliency of the electric distribution grid, which directly impacts the life of citizens and national economy. The proposed control and optimization architectures are flexible, adapt to changing operating scenarios, respond quickly and accurately, provide better scalability and robustness, and safely operate the system even when pushed towards the edges by leveraging massive sensor data, distributed computation, and edge computing. The algorithms and platform will be released open source and royalty-free and the project team will work with industry members and researchers for wider usage of the developed algorithms for other CPNI. Developed artifacts as part of the proposed work will be integrated in existing undergraduate and graduate related courses. Undergraduate students will be engaged in research through supplements and underrepresented and pre-engineering students will be engaged through existing outreach activities at home institutions including Imagine U program and 4-H Teens summer camp programs and the Pacific Northwest Louis Stokes Alliance for Minority Participations. Additionally, project team plans to organize a workshop in the third year to demonstrate the fundamental concepts and applications of the proposed control and optimization architecture to advance CPNI. Developed solutions can be extended for range of applications in multiple CPNIs beyond use cases discussed in the proposed work.

While the proposed control architecture with edge computing offer great potential; coordinating decentralized control and optimization is extremely challenging due to variable network and computational delays, several interleavings of message arrivals, disparate failure modes of components, and cyber security threats leading to several fundamental theoretical problems. Proposed work offers number of novel solutions including (a) adaptive and delay-aware control algorithms, (b) Predictive control and distributed optimization with realistic cyber-physical constraints, (c) threat sharing, data-driven detection and mitigation for cyber security, (d) coordination and management of computing nodes, (e) knowledge learning and sharing. Proposed solutions will be a step towards advancing fundamentals in CPNI and in engineering next generation CPNI. The proposed work also aims to use high fidelity testbed to evaluate developed algorithms and tools for specific CPNI: electric distribution grid.

Safety critical medical systems increasingly aim to incorporate learning-enabled components that are developed using machine learning and AI. While the impact of these learning-enabled medical cyber-physical systems (LE-MCPS) are revolutionizing personalized patient care and health outcomes, assuring their safety and efficacy remains a formidable challenge. Existing model-based design paradigms for learning-enabled cyber-physical systems require an abundance of ?clean? data or high-fidelity simulators ? unfortunately, LE-MCPS do not have that luxury. Consequently, LE-MCPS development strongly depends on experimentation to generate data for design and assurance. The ethical and economic constraints of working in safety-critical medical applications necessitate experimentation efficiency. Yet, experimental design and learning-enabled component design are often weakly coupled -- which contributes to inefficiencies, increased development costs, and increased patient risk. This CAREER proposal aims to develop foundations and tools for assuring learning-enabled medical cyber physical systems (MCPS) by bridging-the-gap between experimentation and model-based design. Specifically, the research focuses on leveraging model-based design techniques to address foundational challenges associated with experimental design (ante-experimentation), protocol execution (during experimentation), and system assurance (post-experimentation). The project?s broader significance will advance the state-of-the-art in medical system design, accelerate learning-enabled CPS (LE-CPS) innovation, and provide abundant interdisciplinary and use-inspired education opportunities and outreach activities.

The goal of this project is to develop foundations and tools for assuring LE-MCPS by bridging-the-gap between experimentation and model-based design. The proposed research will result in a high-assurance LE-CPS design framework spanning ante-, intra-, and post-experimentation. Prior to experimentation, this work will develop foundational techniques to address gaps in traditional experimental designed exposed by high-assurance LE-CPS design. During experimentation, new platforms and capabilities will be realized that can support tamper-evident run-time experimental data curation for assuring LE-CPS. After experimentation, techniques that leverage historical evidence and experimental data will maximally assure LE-CPS designs. Foundations developed in the project are prospectively evaluated in industrial LE-MCPS applications. While the research is motivated by medical scenarios, the developed technologies are immediately applicable to a wide range of LE-CPS applications.

The long-term goal of this project is to build flexible models and efficient algorithms for large-scale, multi-agent, and uncertain cyber-physical systems. In settings such as traffic management, for example, practitioners face fundamental challenges due to complex dynamics, hierarchical influence, noncooperative actors, and hard-to-model uncertainty. Strong simplifying assumptions have become essential: for instance, many theoretical models of road networks take the form of static, deterministic, and/or aggregative games. In these instances, static assumptions make it possible to predict the aggregate impact of decisions such as tolling on traffic patterns. However, neglecting temporal dynamics and feedback effects can lead city planners to make myopic decisions, which may have unintended consequences as drivers adapt to one another's behavior over time. This project develops theoretical and algorithmic techniques to address some of the underlying challenges and will also support mentoring of graduate and undergraduate researchers, development of undergraduate course material, and outreach to local underrepresented communities. This NSF CAREER project aims to develop a sound algorithmic basis for game-theoretic inference and design in dynamic and multi-agent CPS. The specific goals of this project are threefold. The first goal is to formalize and solve a set of structural inference problems in noncooperative games that arise in transportation. For example, one such problem is to discover hierarchies of influence among decision-makers from observations of their actions. The second goal of this project is to design dynamic, time-varying mechanisms which influence agents? decisions and induce desired outcomes. In transportation systems, these mechanisms correspond to tolls, bus routes, timetables, etc. The third and final goal considers stochastic variants of the aforementioned games and aims to develop a computationally-tractable theory of time-varying, feedback decision-making in these settings. This project will enable the analysis and design of cyber-physical systems which interact with one another in complex hierarchies and enable planners and regulators to guide these systems toward desired outcomes. Theory and algorithms will be validated in a physical laboratory testbed which emulates urban driving, via large-scale simulation of traffic in the city of Austin and using French air traffic management data

This NSF project aims to solve the problem of oscillations faced by power grids with high penetration of renewable energy that are disrupting system operations. The project will bring transformative change in fundamental understanding of such phenomena and propose new control strategies to damp the oscillations by leveraging the cyber layer of the power grid including sensors and communication network. This will be achieved by novel approaches of modeling the cyber physical power grid that can capture such phenomena, and centralized and decentralized controls that can damp such oscillations even in presence of anomalies in sensor measurements including cyber-attacks. The intellectual merits of the project include development of computationally manageable cyber physical models, novel sensor grouping and algorithms for data recovery from corruption, and control methods that do not rely on detailed renewable plant models. The broader impacts of the project include solving a major impediment of renewable energy integration that can help tackle climate change, integrating the proposed research in summer camps with high school students, offering summer internships for underrepresented minorities, informing curricula, and student engagement through Penn State?s Center for Engineering Outreach and Inclusion. The proposed project has three key thrusts addressing the sub- and super-synchronous oscillations (SSOs) in presence of inverter-based resources (IBRs). First, a scalable, computationally manageable, and linearizable dynamic phasor-based modeling framework with unbalance simulation capability for grids with high penetration of IBRs is proposed, which is coupled to a realistic cyber layer model with data packet drops and delays. Second, a centralized remedial action scheme based damping control is proposed that relies on three steps ? (a) offline phasor measurement unit (PMU) placement, online dynamic signal grouping, and signal recovery from sparse and non-sparse corruption, (b) detection and source localization of SSOs using dissipating energy flow (DEF) approach, and (c) determination of generation re-dispatch through a novel DEF sensitivity calculation. Third, decentralized modulation-based corruption-resilient SSO damping is proposed based on model predictive control (MPC) framework that does not require knowledge of IBR models, wherein the computational burden will be reduced by estimating a map by offline training of neural networks.

Understanding the root cause of behavior is imperative for informed decision-making and preventing ineffective or biased policies. Currently, most AI-based learning and control modules embedded in cyber-physical systems (CPS) rely on statistical correlation rather than causality for decision-making. This not only results in incorrect decisions but also hinders the interpretability of learning, limiting transferability and scalability. This CAREER proposal aims to bridge the gap between causal inference and the growing capabilities of reinforcement learning (RL) in CPS. The proposed methods are transformative to a wide range of CPS applications, enabling more efficient and effective decision-making processes in autonomous and swarm CPS such as self-driving cars, drones, industrial robots, and swarm robots. This NSF CAREER proposal proposes a set of temporal causal RL and control approaches for CPS by leveraging the reasoning capabilities of temporal logics and causal diagrams in single-agent, multi-agent, and swarm system settings. The tools we develop will be implemented on multiple CPS testbeds and integrated with the proposed education plan. The proposed algorithms have the following unique and innovative features. Firstly, we will develop computationally efficient tools that can discover temporal causal knowledge from both observational and interventional data of a CPS in performing RL to improve the sampling efficiency and transferability. Secondly, we will develop multi-agent RL approaches for CPS in cooperative, non-cooperative, and incomplete information stochastic game environments where temporal causal knowledge is discovered in a distributed way for expediting RL. Lastly, we will develop scalable RL-based control methods for swarm systems utilizing temporal causal reasoning over agent-level features and swarm-level features such as densities and generalized moments. The education plan will impact the next generation of CPS and AI engineers and researchers through AI-assisted adaptive and interactive teaching, temporal-logic-based educational games, online interactive educational website design for temporal causal RL, and workshops and webinars with industrial partners.