The research goal of this project is to design a comprehensive methodology for cybersecurity monitoring and mitigation of the electric power distribution system with a multitude of dynamical devices that are prone to cyberattacks as well as power electronically interfaced renewable generation units. The power electronics interfaces complicate the control and operation of the distribution system because of their sensitivity to undesired disturbances. By promoting a more secure power system, this work helps increase the resiliency of the distribution system to extreme events, which is in line with the U.S. national priorities. Moreover, this work promotes teaching and learning by (i) educating and training graduate students through several programs including a week-long practicum hosted at an industry site; (ii) disseminating the results of our work via conference presentations and publications; and (iii) introducing significant updates in the syllabi of two courses. This work also increases participation of underrepresented minorities via active recruitment.
The overarching research goal of this project is to design a comprehensive methodology for cybersecurity monitoring and mitigation in systems with a multitude of dynamical devices that are prone to cyberattacks. To demonstrate the performance of our proposed algorithms, we study their application for an electric power distribution system as a critical cyberinfrastructure, which includes substations, feeder devices, and smart meters. The electric power grid's growing dependency on information and communication technology (ICT) significantly increases its vulnerability to cyber attacks. A major area of concern is the security of supervisory control and data acquisition (SCADA) systems that the power grids rely upon for monitoring and control functions. To address these challenges, our specific objectives are (i) detection of cyberthreats on the distribution system, (ii) mitigation of and response to cyberintrusions especially with power electronically interfaced renewables via multiagent-based algorithms as well as trajectory shaping and frequency regulation strategies, and (iii) preparing the next generation of cyber-aware engineers.
This Cyber-Physical Systems (CPS) project will develop advanced artificial intelligence and machine-learning (AI/ML) techniques to harness the extensive untapped climatic resources that exist for direct solar heating, natural ventilation, and radiative and evaporative cooling in buildings. Although these mechanisms for building environment conditioning are colloquially termed "passive," their performance depends strongly on the intelligent control of operable elements such as windows and shading, as well as fans in hybrid systems. Towards this goal, this project will create design methodologies for climate- and occupant-responsive strategies that control these operable elements intelligently in coordination with existing building heating ventilation and air conditioning systems, based on sensor measurements of the indoor and outdoor environments, weather and energy forecasts, occupancy, and occupant preferences. The solutions developed in this project can potentially result in substantial reduction in greenhouse gas emissions generated from space heating, cooling, and ventilation. The developed techniques may be particularly valuable in affordable housing by reducing energy costs under normal conditions and improving passive survivability during extreme events and power outages.
Specifically, this project will create intelligent passive and hybrid conditioning systems that optimally leverage climatic resources in the form of temperate outdoor air and sunlight, harness these resources at the building envelope and redistribute them within the building?s microclimates, and learn to respond to changing weather and evolving occupant needs. The project will advance foundational analysis and design tools for a class of physics-informed machine learning models for systems governed by local energy and mass conservation laws. These so-called locally interactive bilinear ?ow models have broad applicability beyond the specific physical building systems studied in this project. From a fundamental cyber physical systems standpoint, the researchers will establish analytical certificates for learning and control algorithms designed for this class of systems, bridging the gap between purely data-driven strategies and physics-based models. Finally, the project will provide a systematic mechanism to evaluate climate resources available through the intelligent operation of passive systems, bridging a key gap in current understanding. Demonstrations in occupied buildings will provide key insights and evidence to support the applicability of the researched tools in the real world. This effort will also develop and present educational modules to attract middle and high school students to encourage careers in sustainable engineering through the RPI Engineering Ambassadors program; at the same time, project outcomes will also support community engagement with science and technology through the University of Oregon Sustainable City Year program.
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.
Engineering requirements often make incompatible demands on models. Detailed models make highly accurate predictions, but coarse models are easier to interpret. This project will develop techniques to overcome this inherent contradiction. On the one hand, data science and machine learning techniques allow us to efficiently construct black box predictive models with limited generalizability. At the same time, recent advances in information geometry have produced model reduction methods that systematically derive simple, interpretable models from physical first principles that summarize relevant mechanisms needed for model transferability. Combining these technologies will enable useful mappings between ?physically explainable? reduced models and quantitative data. These data-driven tools will enable ?the best of both worlds? ? physically interpretable models that make quantitative predictions. We will combine a meaningful, qualitatively correct but quantitatively inaccurate reduced model with a data-driven transformation. The project team brings together domain-specific expertise in physical modeling, energy systems, and data-driven learning. We will apply this approach to address key operational challenges in interconnected energy networks. The enabling technology will apply to modeling any complex cyber-physical system.
The application of acoustic monitoring in ecological sciences has grown exponentially in the last two decades. It has been used to answer many questions, including detecting the presence or absence of animal species in an environment, evaluating animal behavior, and identifying ecological stressors and illegal activities. However, current uses are limited to the coverage of relatively small geographic areas with a fixed number of sensors. Animal-borne GPS-based location trackers paired with other sensors are another widely used tool in aiding wildlife conservation and ecosystem monitoring. Since capturing and collaring wild animals is a traumatic event for them, as well as being expensive and resource-intensive, multiyear deployments are required. There are severely limited opportunities to recharge batteries making relatively power-hungry sensing, such as acoustic monitoring, out of reach for existing tracking collars. The aim of the A3EM project is to devise an animal-borne adaptive acoustic monitoring system to enable long-term, real-time observation of the environment and behavior of wildlife. Animal-borne acoustic monitoring will be a novel tool that may provide new insights into biodiversity loss, a severe but underappreciated problem of our time. Combining acoustic monitoring with location tracking collars will enable entirely new applications that will facilitate census gathering and monitoring of threatened and endangered species, detecting poachers of elephants in Africa or caribou in Alaska, and evaluating the effects of mining and logging on wildlife, among many others. All data, hardware designs, and software source code will be released to the public domain, enabling tracking collar manufacturers to include the technology within their products. A3EM constitutes a complex cyber-physical architecture involving humans, animals, distributed sensing devices, intelligent environmental monitoring agents, and limited power and network connectivity. This intermittently connected CPS, with a power budget an order of magnitude lower than typical, calls for novel approaches with a high level of autonomy and adaptation to the physical environment. A3EM will employ a unique combination of supervised and semi-supervised embedded machine learning to identify new and unexplored event classes in a given environment, dynamically control and adjust parameters related to data acquisition and storage, opportunistically share knowledge and data between distributed sensing devices, and optimize the management of storage and communication to minimize resource needs. These methods will be evaluated through the creation of a wearable acoustic monitoring system used to support ecological applications such as enhanced wildlife protection, rare species identification, and human impact studies on animal behavior.
The overarching goal of the proposed research is to derive critical information and characterization of large scale generic nonlinear dynamical systems using limited observables. In the present state-of-the-art in data-driven dynamical system analysis, all the underlying state measurements and the time evolution of these states are required. Access to all of the dynamical states measurements in real-world is impossible or expensive. The objective of the proposal is to develop data-driven tools for dynamic system identification, classification and root-cause analysis of dynamic events, and prediction of system evolution. The research team will specifically conduct research on using available measurements to perform near real-time applications for various dynamic events that occur in electric power systems. The data analytics proposed are applicable to general non-linear dynamic systems and can be easily applied to other cyberphysical systems (CPS). More broadly, there is a large effort in the CPS and control community to model real world systems that we all interact with on a daily basis (such as transportation systems, communication networks, world wide web, etc.) as dynamical systems and thus, the theory and techniques developed through this project will enable online monitoring of these critical systems, allowing operators to quickly analyze these systems for any unstable/anomalous behavior from minimal data streams. The project will promote various educational and outreach activities including developing new courses, short courses, activities in schools, and scholarships for women and underrepresented minority students.
Overview: The goal of this proposal is to develop operator theoretic data analytics techniques for dynamic systems with limited measurements to identify the underlying non-linear dynamical system and characterize their behavior such as causal interactions between constituent components, stability monitoring, identifying targets for control. The proposed research is in the domain of "Technology for cyber-physical systems". The novelty of the proposed methods is that they do not require the dynamic states but can utilize system outputs, making it applicable to real-world dynamical systems. Power systems are rapidly evolving with increased deployment of sensors like the phasor measurement units (PMUs) that have high accuracy and high sampling frequencies (up to 120 Hz). These measurements will be used to develop an equivalent linear representation in a higher dimensional function space that can be used for online identification and characterization of nonlinear dynamics of the power grid. Further, machine learning techniques will be formulated to learn effective dictionary functions for the scalable deployment of proposed method. Using the proposed system identification method, the project will develop the theory and methodology for data-driven Information Transfer based causality mapping for detection and localization of system stress and dynamic coupling between the systems components. Specific applications for power grids will include stability monitoring, trajectory prediction and identification of targets for controlling adverse dynamic behavior. The methods are evaluated by an integrated power-cyber co-simulator (IPCC) that integrates power transmission, distribution and communication systems to generate synthetic sensor data for large systems under various dynamic scenarios. The IPCC will be able to model intermediate communication networks that cause measurement inconsistencies like delays, packet drop, etc. The Iowa State University's hardware in the loop cyber-physical testbed will be used to validate and evaluate some of the online applications like stability monitoring and trajectory prediction for large power grid topologies.
Smart home products have become extremely popular with consumers due to the convenience offered through home automation. In bridging the cyber-physical gap, however, home automation brings a widening of the cyber attack surface of the home. Research towards analyzing and preventing security and safety failures in a smart home faces a fundamental obstacle in practice: the poor characterization of home automation usage. That is, without the knowledge of how users automate their homes, it is difficult to address several critical challenges in designing and analyzing security systems, potentially rendering solutions ineffective in actual deployments. This project aims to bridge this gap, and provide researchers, end-users, and system designers with the means to collect, generate, and analyze realistic examples of home automation usage. This approach builds upon a unique characteristic of emerging smart home platforms: the presence of "user-driven" automation in the form of trigger-action programs that users configure via platform-provided user interfaces. In particular, this project devises methods to capture and model such user-driven home automation to generate statistically significant and useful usage scenarios. The techniques that will be developed during the course of this project will allow researchers and practitioners to analyze various security, safety and privacy properties of the cyber-physical systems that comprise modern smart homes, ultimately leading to deployments of smart home Internet of Things (IoT) devices that are more secure. The project will also produce and disseminate educational materials on best practices for developing secure software with an emphasis on IoT devices, suitable for integration into existing computer literacy courses at all levels of education. In addition, the project will focus on recruiting and retaining computer science students from traditionally underrepresented categories.
This project is centered on three specific goals. First, it will develop novel data collection strategies that allow end-users to easily specify routines in a flexible manner, as well as techniques based on Natural language Processing (NLP) for automatically processing and transforming the data into a format suitable for modeling. Second, it will introduce approaches for transforming routines into realistic home automation event sequences, understanding their latent properties and modeling them using well-understood language modeling techniques. Third, it will contextualize the smart home usage models to make predictions that cater to security analyses specifically and develop tools that allow for the inspection of a smart home?s state alongside the execution of predicted event sequences on real products. The techniques and models developed during the course of this project will be validated with industry partners and are expected to become instrumental for developers and researchers to understand security and privacy properties of smart homes.
Propelled by the growth in demand for artificial intelligence-enabled applications, the past decade has witnessed the emergence of Collaborative Cyber-Physical Learning Systems (CCPLS). CCPLS carry out distributed, learning-based processing tasks through coordination among Cyber-Physical System (CPS) devices, and are envisioned to provide critical functionality across the commercial and defense sectors in the next several years. However, the data generated by CCPLS is often large-scale, high-dimensional, heterogeneous, and time-varying, which poses critical challenges for intelligence modeling. Concurrently, unmanned vehicles, in particular Unmanned Aerial Vehicles (UAVs), have shown promise of scaling up information-sharing in CCPLS, especially in under-served regions such as rural areas. The project's novelties are in establishing a concrete foundation for UAV-CCPLS integration that unifies the associated learning, networking, and communication design aspects around appropriate intelligence metrics. The project's impacts are the development of UAV-assisted CCPLS for smart agriculture tasks, as well as advancing the manufacturing of UAVs and other unmanned vehicles tailored for CCPLS. Project outcomes will be disseminated by releasing open-source software and research videos and tutorials. The investigators will further engage in Curriculum development, diversity, and outreach activities including mentoring undergraduate researchers. Research investigations center around three interconnected thrusts. Thrust 1 develops a novel UAV-assisted intelligence framework for CCPLS and introduces a precise, task-oriented notion of data dynamics and heterogeneity. Additionally, this thrust develops a `learning for learning? framework that aims to predictively estimate the rate of data dynamics. Thrust 2 investigates methodologies for jointly optimizing resource utilization and intelligence quality through co-design of UAV trajectories and UAV-to-CPS network establishment. The data dynamics framework from Thrust 1 is integrated into this design through an online, network-aware sequential decision-making framework. Finally, Thrust 3 develops CCPLS communication protocols based on learning-aware uplink and downlink wireless beamformers and over-the-air aggregation methods. These protocols are tailored to the specific needs of the UAV-assisted learning systems, e.g., the transmission of noisy information over UAV-to-UAV and UAV-to-access point communication links. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.
This research project will study novel methods for designing sequential, non-myopic, online decision procedures for societal-scale cyber-physical systems such as public transit, emergency response systems, and power grid, forming the critical infrastructure of our communities. Online Optimization of these systems entails taking actions that consider the tightly integrated spatial, temporal, and human dimensions while accounting for uncertainty caused due to changes in the system and the environment. For example, emergency response management systems (ERM) operators must optimally dispatch ambulances and help trucks to respond to incidents while accounting for traffic pattern changes and road closures. Similarly, public transportation agencies operating electric vehicles must manage and schedule the vehicles considering the expected travel demand while deciding on charging schedules considering the overall grid load. The project's proposed approach focuses on designing a modular and reusable online decision-making pipeline that combines the advantages of online planning methods, such as Monte-Carlo Tree Search, with offline policy learning methods, such as reinforcement learning, promising to provide faster convergence and robustness to changes in the environment. The research activities of the proposed project are complemented by educational activities focusing on designing cloud-based teaching environments that can help students and operators with prerequisite domain and statistical knowledge to design, manage, and experiment with decision procedures.
The societal-scale CPS that we study have spatial-temporal properties. The spatial aspect refers to the location-specific state variables such as traffic congestion, transportation demand, and the frequency with which incidents occur at a location. The temporal aspect refers to the dynamic nature of these systems---traffic congestion evolves over time. Non-myopic decisions entail selecting actions over time under uncertainty while accounting for future impact and demand for resources. The combined research and education efforts proposed in the project focus on answering the following critical questions for these systems - first, how do we solve the challenge of sampling future state/ environmental actions across a high-dimensional space while also tackling the challenge of non-stationarity? Second, how do we address the need for robust, fast non-myopic planning that also tackles potential non-stationarity? And third, how do we make it possible to engage non-computer science students and community partners with the solutions built using approaches pioneered in the project? The proposed approach involves investigating novel machine learning methods, such as normalizing flows for designing generative models and an innovative approach to design planning algorithms using a policy-augmented hybrid Monte-Carlo Tree Search approach. A significant effort of the project will focus on complementing fundamental research with the design of a cloud-based visual domain-specific modeling environment that can help explain the design, operation, and introspection of methods by using a block-based compositional approach. The work will be augmented with course modules and online tutorials accompanying the cloud-based environment.
This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.