Deep Neural Networks (DNN) enabled Cyber-Physical Systems (CPS) hold great promise for revolutionizing many industries, such as drones and self-driving cars. However, the current generation of DNN cannot provide analyzable behaviors and verifiable properties that are necessary for safety assurance. This critical flaw in purely data-driven DNN sometimes leads to catastrophic consequences, such as vehicle crashes linked to self-driving and driver-assistance technologies. On the other hand, physics-model-based engineering methods provide analyzable behaviors and verifiable properties, but do not match the performance of DNN systems. These considerations motivate the work in this project which aims at physics-model-based neural networks (NN) redesign, yielding HyPhy-DNN: a hybrid self-correcting physics-enhanced DNN framework. HyPhy-DNN will provide the performance of DNNs and the analyzability and verifiability of physical models, thus providing a foundation for verifiably safe self-driving cars, drones, and other CPS systems. Moreover, the HyPhy-DNN will fundamentally advance the integration of deep learning and robust control to enable safety- and time-critical CPS to safely operate with high performance in unforeseen and dynamic environments.

This NSF CPS project aims to develop new techniques for modeling cyber-physical systems that will address fundamental challenges associated with scale and complexity in modern engineering. The project will transform human interaction with complex cyber-physical and engineered systems, including critical infrastructure such as interconnected energy networks. This will be achieved through a novel combination of data-driven techniques and physics-based approaches to give mathematical and computational models that are at once abstract enough to be understood by humans making key engineering decisions and precise enough to make quantitative predictions. The intellectual merits of the project include a novel confluence of emerging data science and model-analysis methods, including manifold learning and information geometry. The broader impacts of the project include the training of undergraduates, including those from underrepresented communities, several outreach activities, and publicly available open-source software.

Enhancing resource availability, health, security, and a sense of well-being can be enhanced by our ability to sense, analyze, and act on our world with efficient, safe, and secure engineered systems. To realize such systems requires a deep understanding of the interfaces between the cyber and physical worlds, leading to the establishment of the field of Cyber Physical Systems (CPS). While CPS as a discipline and application-enabler has evolved tremendously over the past decade, current graduate training does not sufficiently prepare students for fundamental discovery and innovation in CPS nor for effective translation of research to application development. This National Science Foundation Research Traineeship (NRT) award to the University of Virginia (UVA) will address this need by training graduate students to pursue fundamental CPS discovery and innovation and to collaborate with application domain experts to realize a smarter planet, specifically in the areas of smart health, smart cities, and autonomous systems. The traineeship anticipates training one hundred fifty-eight (158) master's and doctoral students, including thirty-three (33) funded PhD trainees, from civil and environmental engineering, computer science, electrical and computer engineering, mechanical and aerospace engineering, and systems and information engineering. Current CPS graduate training is lacking in three critical ways that will be addressed in this project. First, current courses do not instill the integrative knowledge needed for new scientific discovery and translational applications in the field of CPS. Second, most students do not have a sufficiently robust experience of convergence activities as part of their training. This lack exists not only in traditional engineering and computing education but also extends into analyses of target application domains and associated grand challenges. Third, explicit professional development is absent in most graduate training. Such exposure is critical in CPS given the field's potent role in our ever-evolving smart world, and CPS practitioners must attend to social issues regarding ethics, safety, privacy, communication, and policy. This traineeship will address these issues and drive CPS graduate education nationwide by developing a novel, comprehensive graduate training program that involves orientation, normalization modules, elective courses, experiential convergence research activities, professional development workshops, and a hands-on, testbed-driven educational curriculum. Students will acquire both the technical depth and the integrative transdisciplinary understanding of CPS and its associated application domains to be successful in CPS-related careers. During their time in the program, trainees will engage in transdisciplinary CPS research on projects related to challenges in smart health, smart cities, and autonomous systems. These research efforts have the potential for significant scientific and application impact, such that a smarter world can be achieved and associated societal grand challenges can be addressed. The NSF Research Traineeship (NRT) Program is designed to encourage the development and implementation of bold, new potentially transformative models for STEM graduate education training. The program is dedicated to effective training of STEM graduate students in high priority interdisciplinary research areas through comprehensive traineeship models that are innovative, evidence-based, and aligned with changing workforce and research needs.

This award is funded in whole or in part under the American Rescue Plan Act of 2021 (Public Law 117-2). This project will develop novel modeling, control, and optimization methods for connected and automated vehicles to operate in human-dominated traffic to improve the efficiency and sustainability of the urban transportation system while respecting individual drivers? unique behaviors and social norms accordingly. The significance of the research is highlighted by the following two needs. First, the inefficiency of the urban transportation system has resulted in substantial fuel waste and emissions over the decades. Leveraging vehicles? growing autonomy and connectivity, a significant boost of energy efficiency, emission performance, and traffic management can be achieved through dedicated control and optimization of vehicle maneuvers and routes. Second, human drivers will remain the majority of operators on the road in the foreseeable future. The resulting mixed traffic where connected and automated vehicles and human drivers share the road with frequent interactions requires detailed modeling of human drivers? behaviors in a socially compatible context. The proposed research can generate socioeconomic incentives such as improving the efficiency of the urban transportation system and promoting technology acceptance for sustainable mobility, thereby alleviating the nation's energetic and environmental concerns. The scientific outcome of the project will advance convergent research areas of control theory, optimization, human behavioral study, and machine learning. The project will involve an interdisciplinary team of students through hands-on research opportunities at Texas Tech University, which has been historically and actively engaged in serving the traditionally underrepresented student body in STEM, contributing towards equitable and inclusive educational and social outcomes.

Robotic manipulation and automation systems have received a lot of attention in the past few years and have demonstrated promising performance in various applications spanning smart manufacturing, remote surgery, and home automation. These advances have been partly due to advanced perception capabilities (using vision and haptics) and new learning models and algorithms for manipulation and control. However, state-of-the-art cyber-physical systems remain limited in their sensing and perception to a direct line of sight and direct contact with the objects they need to perceive. The goal of this project is to design, build, and evaluate a cyber-physical system that can sense, perceive, learn, and manipulate far beyond what is feasible using existing systems. To do so, the research will explore the terahertz band, which offers a new sensing dimension by inferring the inherent material properties of objects via wireless terahertz signals and without direct contact. This project will also explore radio-frequency signals that can traverse occlusions. Building on these emerging sensing modalities, the core of the project focuses on developing full-spectrum perception, control, learning, and manipulation tasks. The success of this project will result in CPS system architectures with unprecedented capabilities, enabling fundamentally new opportunities to make robotic manipulation more efficient and allowing robots to perform new complex tasks that have not been possible before.

This project provides a workshop ?Cyber Physical Systems (CPS) Rising Stars? to be held at the University of Virginia. The purpose of the workshop is to identify and mentor outstanding Ph.D. students and postdocs who are interested in pursuing academic careers in CPS areas. This workshop gives participants a chance to gain insights about navigating the early stages of careers in academia, as well as provide networking opportunities with faculty and peers, opening the door for on-going collaboration and professional support for years to come. The workshop includes a program committee comprised of faculty with extensive background in CPS who will select students for the workshop. One workshop aims to increase representation and diversity in CPS, by encouraging applications from women,underrepresented minorities (URM), and persons with disabilities. A specific effort is also being made to recruit workshop participants who may be at non-R1 schools. This will be the second CPS Rising Stars Workshop. This workshop will be modeled after the first including a variety of sessions designed to build a cadre of researchers in CPS and foster future collaborations amongst the students in the field. The first Rising Star workshop was successful in recruiting a high percentage of URM students. This workshop will retain this recruiting focus in addition to paying special attention to selection of non-R1 candidates.

Connected Automated Vehicle (CAV) applications are expected to transform the transportation landscape and address some of the pressing safety and efficiency issues. While advances in communication and computing technologies enable the concept of CAVs, the coupling of application, control and communication components of such systems and interference from human actors, pose significant challenges to designing systems that are safe and reliable beyond prototype environments. Realizing CAV applications, in particular in situations where humans may partly remain in the loop, requires addressing uncertainties that arise from human input. Large scale deployment of CAVs will also require addressing challenges in coordination of actions among CAVs and with human operated systems. To address these challenges, this project develops a novel model-based stochastic hybrid systems (SHS)-theoretic approach that relies on describing and communicating behaviors of actors in the system in the form of evolving SHS using Bayesian learning. The models are then utilized in a stochastic model predictive control (SMPC) framework for optimal coordination of actions. The proposed research will provide wide-ranging societal benefits through three major impact areas: first, by advancing research in stochastic communication-aware control design for hybrid systems; second, through the development of new models and advanced controllers to address the emerging challenges of coordinating mixed systems of automated and manned vehicles, hence opening new vistas in other areas involving general multi-agent systems; and third, through educational and outreach activities that are natural extensions of this multidisciplinary research. This project is also the first fruits of a recent National Science Foundation/Deutsche Forschungs Gesellschaft (NSF/DFG) collaboration on cyber-physical systems (CPS). Through this collaboration, NSF funds the US component (University of Central Florida and University of Georgia) while the German partners (University of Technology and University of Koblenz-Landau) are funded by DFG.

Large-scale systems with societal relevance, such as power generation systems, are increasingly able to leverage new technologies to mitigate their environmental impact, e.g., by harvesting energy from renewable sources. This NSF CPS project aims to investigate methods and computational tools to design a new user-centric paradigm for energy apportionment and distribution and, more broadly, for trustworthy utility services. In this paradigm, distributed networked systems will assist the end users of electricity in scheduling and apportioning their consumption. Further, they will enable local and national utility managers to optimize the use of green energy sources while mitigating the effects of intermittence, promote fairness, equity, and affordability. This project pursues a tractable approach to address the challenges of modeling and designing these large-scale, mixed-autonomy, multi-agent CPSs. The intellectual merits include new scalable methods, algorithms, and tools for the design of distributed decision-making strategies and system architectures that can assist the end users in meeting their goals while guaranteeing compliance with the fairness, reliability, and physical constraints of the design. The broader impacts include enabling the automated design of distributed CPSs that coordinate their decision-making in many applications, from robotic swarms to smart manufacturing and smart cities. The research outcomes will also be used in K-12 and undergraduate STEM outreach efforts.

Machine learning has found many successes in modern commercial application domains like computer vision, speech analysis, and natural language processing. However, its broader use in critical infrastructure cyber-physical systems (CI-CPS), such as, energy, water, transportation, and oil and natural gas systems, has been far less than ideal. This is mainly due to concerns with the reliability of existing machine learning techniques and the lack of explainability of the learned models. Moreover, CI-CPS often borrow techniques directly from commercial applications that fail to consider physical and topological constraints inherent in these systems. Security of machine learning has been extensively studied recently, revealing vulnerabilities of machine learning models and the effectiveness in deviating learning outcomes by polluting the model input. This is especially devastating in CI-CPS where learning is used for safety-critical operations and such deviation can cause irreversible harm to people and physical assets. Secure machine learning that models unique CI-CPS constraints is thus a much needed research area and is the focus of this project. This proposal intersects three fields - security, machine learning, and CI-CPS - to enhance the safety and resiliency of essential infrastructures in modern society. We use two CI-CPS, power systems and transportation systems, as target application domains to illustrate the general applicability of the proposed approach. The proposed work is carried out by four research tasks. First, the project will devise a suitable threat model under which adversarial machine learning attacks, ConAML, are developed subject to CI-CPS constraints. Second, the project will propose a mitigation method for ConAML attacks by introducing random input padding in both training and inference. Third, the project will propose a new ?data-representation-model-task? association framework that realizes secure constrained machine learning from ground up, by designing a variation Dirichlet-network that bridges the input data with machine learning models in the representation space instead of the raw data space. Lastly, the project team will apply the proposed secure constrained machine learning to electric load forecasting and traffic forecasting, implement these applications in testbeds, and evaluate their security and performance under ConAML attacks. The proposed research seeks to improve the security, reliability and resiliency of CI-CPS. It contributes to the knowledge base of secure machine learning for CI-CPS, and applies to all safety-critical large interconnected CPS. The multi-disciplinary nature of the proposed work lends itself to cross-disciplinary education and training of future scientists and engineers.

Cognitive workload refers to the level of mental effort put forth by an individual in response to a cognitive task. Unfortunately, no technology currently exists that can monitor an individual?s levels of cognitive workload in real-world environments using a seamless, reliable, and low-cost approach. We propose to fill this gap by using a novel magnetocardiography (MCG) system worn upon the subject?s chest to allow the sensor to collect the magnetic fields that are naturally emanated by the heart and associated with brain activity. This science is anticipated to greatly accelerate progress in such diverse disciplines as pediatric concussion recovery, pilot training, improved user-machine interfaces, injury prevention in construction environments, increased human performance in risky missions, and improved education outcomes. In addition to advances in basic science, the proposed research is expected to be of significant interest to students and the public. Through targeting interdisciplinary education and diverse recruitment, we intend to expose new audiences to STEM concepts via workshops and family-friendly outings. The proposed MCG sensor is smartly integrated in a Cyber-Physical System (CPS) with two inter-connected loops: (a) a human-in-the-loop that addresses changes in the thresholds of different cognitive states as a function of time, and (b) a non-human-in-the-loop that adapts the system?s algorithmic and hardware components for high-accuracy classification of cognitive workload with minimum resource usage. Our goals are to: (1) Build a knowledgebase concerning the impact of hardware/algorithmic advances upon MCG sensor performance in real-world settings. (2) Explore the classification of cognitive workload from MCG data and close the loop with the wearer for dynamic calibrations that address the time-varying thresholds of cognitive states. (3) Ensure operability in dynamic real-world settings and close the loop between the cyber and physical sides for minimal resource usage. (4) Validate the CPS within the framework of measuring cognitive workload for children with concussion. Without loss of generality, we select this population given the immense clinical potential: the effects of cognitive activity on pediatric concussion recovery are currently unknown, largely due to the difficulties in quantifying cognitive activity workload.