As the number of low cost computing devices at the edge of network increases, there are greater opportunities to enable novel, innovative capabilities, especially in decentralized cyber-physical systems. For example, a set of networked, collaborating processors at the edge can be used to dynamically detect traffic densities via image processing and then use those densities to control the traffic flow by coordinating traffic light sequences; in a decentralized architecture. In this paper we describe a testbed and an application framework for such applications. Furthermore, we describe a queuing theory-based model for analyzing and optimizing workload placement across the fog nodes and available cloud resources.

Distributed real-time and embedded (DRE) systems executing mixed criticality task sets are increasingly being deployed in mobile and embedded cloud computing platforms, including space applications. These DRE systems must not only operate over a range of temporal and spatial scales, but also require stringent assurances for secure interactions between the system’s tasks without violating their individual timing constraints. To address these challenges, this paper describes a novel distributed operating system focusing on the scheduler design to support the mixed criticality task sets. Empirical results from experiments involving a case study of a cluster of satellites emulated in a laboratory testbed validate our claims.

Businesses today are moving their infrastructure to the cloud environment to reduce their IT budgets and improve compliance to regulatory control. By using the cloud, industries also aim to deploy and deliver new applications and services rapidly with the ability to scale their applications horizontally and vertically to meet customer demands. Despite these trends, reliance on old school IT management and administration has left a legacy of poor manageability and inflexible infrastructure. In this realm, the DevOps community has made available various tools for deployment, management, and orchestration of complex, distributed cloud applications. Despite these modern trends, the continued reliance on old school IT management and administration methods have left a majority of developers lacking with the domain expertise to create, provision and manage complex IT environments using abstracted, script-centric approaches provided by DevOps tools. To address these challenges while emphasizing vendor-agnostic approaches for broader applicability, we present a model-driven generative approach that is compliant with the TOSCA specification where the users can model their business-relevant models, requiring only little domain expertise. In this context, the paper describes the metamodel of the domain-specific language that abstracts the business-relevant application requirements to generate the corresponding fully deployable infrastructure-as-code solutions, which can deploy, run, and manage the applications in the cloud environment. Our work is focused on realizing a high-performance deployment and management platform for cloud applications with an emphasis on extensibility, (re)usability, and scalability. We validate our approach by a prototypical application model and present a user study to evaluate its relevance.

An emerging trend in Internet of Things (IoT) applications is to move the computation (cyber) closer to the source of the data (physical). This paradigm is often referred to as edge computing. If edge resources are pooled together they can be used as decentralized shared resources for IoT applications, providing increased capacity to scale up computations and minimize end-to-end latency. Managing applications on these edge resources is hard, however, due to their remote, distributed, and possibly dynamic nature, which necessitates autonomous management mechanisms that facilitate application deployment, failure avoidance, failure management, and incremental updates. To address these needs, we present CHARIOT, which is an orchestration middleware capable of autonomously managing IoT systems consisting of edge resources and applications. CHARIOT implements a three-layer architecture. The topmost layer comprises a system description language; the middle layer comprises a persistent data storage layer and the corresponding schema to store system information; and the bottom layer comprises a management engine, which uses information stored persistently to formulate constraints that encode system properties and requirements thereby enabling the use of Satisfiability Modulo Theories (SMT) solvers to compute optimal system (re)configurations dynamically at runtime. This paper describes the structure and functionality of CHARIOT and evaluates its efficacy as the basis for a smart parking system case study that uses sensors to manage parking spaces.

Evaluation of key performance indicators (KPIs) such as energy consumption is essential for decision- making during the design and operation of smart manufacturing systems. The measurements of KPIs are strongly affected by several uncertainty sources such as input material uncertainty, the inherent variability in the manufacturing process, model uncertainty and the uncertainty in the sensor measurements of operational data. A comprehensive understanding of the uncertainty sources and their effect on the KPIs is required to make the manufacturing processes more efficient. Towards this objective, this paper proposes an automated methodology to generate a Hierarchical Bayesian network (HBN) for a manufacturing system from semantic system models, physics-based models and available data in an automated manner, which can be used to perform uncertainty quantification (UQ) analysis. The semantic system model, which is a high-level model describing the system along with its parameters is assumed to be available in the Generic Modeling Environment (GME) platform. Apart from semantic description, physics-based models, if available, are assumed to be available in model libraries. The proposed methodology is divided into two tasks – (1) Automated Hierarchical Bayesian network construction using semantic system model, available models and data, and (2) Automated uncertainty quantification (UQ) analysis. A metamodel of a HBN is developed using the GME, along with a syntax representation for the associated conditional probability tables/distributions. The constructed HBN corresponding to a system is represented as an instance model of the HBN metamodel. On the metamodel, a model interpreter is written to be able to carry out the UQ analysis in an automated manner for any HBN instance model conforming to the HBN metamodel. The proposed methodologies are demonstrated using an injection molding process.

We propose a method for verifying persistence of nonlinear hybrid systems. Given some system and an initial set of states, the method can guarantee that system trajectories always eventually evolve into some specified target subset of the states of one of the discrete modes of the system, and always remain within this target region. The method also computes a time-bound within which the target region is always reached. The approach combines flowpipe computation with deductive reasoning about invariants and is more general than each technique alone. We illustrate the method with a case study concerning showing that potentially destructive stick-slip oscillations of an oil-well drill eventually die away for a certain choice of drill control parameters. The case study demonstrates how just using flow-pipes or just reasoning about invariants alone can be insufficient. The case study also nicely shows the richness of systems that the method can handle: the case study features a mode with non-polynomial (nonlinear) ODEs and we manage to prove the persistence property with the aid of an automatic prover specifically designed for handling transcendental functions.

The on-time arrival performance of vehicles at stops is a critical metric for both riders and city planners to evaluate the reliability of a transit system. However, it is a non-trivial task for transit agencies to adjust the existing bus schedule to optimize the on-time performance for the future. For example, severe weather conditions and special events in the city could slow down traffic and cause bus delay. Furthermore, the delay of previous trips may affect the initial departure time of consecutive trips and generate accumulated delay. In this paper, we formulate the problem as a single-objective optimization task with constraints and propose a greedy algorithm and a genetic algorithm to generate bus schedules at timepoints that improves the bus ontime performance at timepoints which is indicated by whether the arrival delay is within the desired range. We use the Nashville bus system as a case study and simulate the optimization performance using historical data. The comparative analysis of the results identifies that delay patterns change over time and reveals the efficiency of the greedy and genetic algorithms.

The emerging Fog Computing paradigm provides an additional computational layer that enables new capabilities in real-time data-driven applications. This is especially interesting in the domain of Smart Grid as the boundaries between traditional generation, distribution, and consumer roles are blurring. This is a reflection of the ongoing trend of intelligence distribution in Smart Systems. In this paper, we briefly describe a component-based decentralized software platform called Resilient Information Architecture Platform for Smart Systems (RIAPS) which provides an infrastructure for such systems. We briefly describe some initial applications built using this platform. Then, we focus on the design and integration choices for a resilient Discovery Manager service that is a critical component of this infrastructure. The service allows applications to discover each other, work collaboratively, and ensure the stability of the Smart System.

As the number of low cost computing devices at the edge of communication network increase, there are greater opportunities to enable innovative capabilities, especially in cyber-physical systems. For example, micro-grid power systems can make use of computing capabilities at the edge of a Smart Grid to provide more robust and decentralized control. However, the downside to distributing intelligence to the edge away from the controlled environment of the data centers is the increased risk of failures. The paper introduces a framework for handling these challenges. The contribution of this framework is to support strategies to (a) tolerate the transient faults as they appear due to network fluctuations or node failures, and to (b) systematically reconfigure the application if the faults persist.

Emerging smart services, such as indoor smart parking or patient monitoring and tracking in hospitals, incur a significant technical roadblock stemming primarily from a lack of cost-effective and easily deployable localization framework that impedes their widespread deployment.To address this concern, in this paper we present a low-cost, indoor localization and navigation system, which performs continuous and real-time processing of Bluetooth low Energy (BLE) and IEEE 802.15.4a compliant Ultra-wideband(UWB) sensor data to localize and navigate the concerned entity to its desired location. To keep deployment costs down, the indoor space in our solution is instrumented with (battery) as well as wired Edison devices, which provide both compute and BLE capabilities. Entities with managerial responsibilities in these scenarios can be equipped with both localization modalities: UWB tags and a BLE capable device (current generation smartphone or tablet), and are set up to maintain the BLE Received Signal Strength Intensity (RSSI) fingerprint map using the UWB positioning data as ground truth. The remaining entities rely exclusively on BLE RSSI fingerprinting-based localization using their smartphones.

Cyber-physical systems (CPS) are smart systems that include engineered interacting networks of physical and computational components. The tight integration of a wide range of heterogeneous components enables new functionality and quality of life improvements in critical infrastructures such as smart cities, intelligent buildings, and smart energy systems. One approach to study CPS uses both simulations and hardware-in-the-loop (HIL) to test the physical dynamics of hardware in a controlled environment. However, because CPS experiment design may involve domain experts from multiple disciplines who use different simulation tool suites, it can be a challenge to integrate the heterogeneous simulation languages and hardware interfaces into a single experiment. The National Institute of Standards and Technology (NIST) is working on the development of a universal CPS environment for federation (UCEF) that can be used to design and run experiments that incorporate heterogeneous physical and computational resources over a wide geographic area. This development environment uses the High Level Architecture (HLA), which the Department of Defense has advocated for co-simulation in the field of distributed simulations, to enable communication between hardware and different simulation languages such as Simulink and LabVIEW. This paper provides an overview of UCEF and motivates how the environment could be used to develop energy experiments using an illustrative example of an emulated heat pump system.

With increasing advances in Internet-enabled devices, large cyber-physical systems (CPS) are being realized by integrating several sub-systems together. Analyzing and reasoning different properties of such CPS requires co-simulations by composing individual and heterogeneous simulators, each of which addresses only certain aspects of the CPS. Often these co-simulations are realized as point solutions or composed in an ad hoc manner, which makes it hard to reuse, maintain and evolve these co-simulations. Although our prior work on a model-based framework called Command and Control Wind Tunnel (C2WT) supports distributed co-simulations, many challenges remain unresolved. For instance, evaluating these complex CPSs require large amount of computational and I/O resources for which the cloud is an attractive option but there is a general lack of scientific approaches to deploy co-simulations in the cloud. Specifically, the key challenges include (i) rapid provisioning and de-provisioning of experimental resources in cloud for different co-simulation workloads, (ii) simulating incompatibility and resource violations, (iii) reliable execution of co-simulation experiments, and (iv) reproducible experiments. Our solution builds upon the C2WT heterogeneous simulation integration technology and leverages Linux container mechanism to provide an integrated tool-suite for specifying experiment and resource requirements, and deploying repeatable cloud-scale experiments. In this work, we present the core concepts and architecture of our framework, and provide a summary of our current work in addressing these challenges.

Reliable operation of electrical power systems in the presence of multiple critical N - k contingencies is an important challenge for the system operators. Identifying all the possible N - k critical contingencies to design effective mitigation strategies is computationally infeasible due to the combinatorial explosion of the search space. This paper describes two heuristic algorithms based on the iterative pruning of the candidate contingency set to effectively and efficiently identify all the critical N - k contingencies resulting in system failure. These algorithms are applied to the standard IEEE-14 bus system, IEEE-39 bus system, and IEEE-57 bus system to identify multiple critical N-k contingencies. The algorithms are able to capture all the possible critical N - k contingencies (where 1 < k < 9) without missing any dangerous contingency.

This paper develops a model-based framework for the quantification and propagation of multiple uncertainty sources affecting the performance of a smart system. A smart system, in general, performs sensing, control and actuation for proper functioning of a physical subsystem (also referred to as a plant). With strong feedback coupling between several subsystems, the uncertainty in the quantities of interest (QoI) amplifies over time. The coupling in a generic smart system occurs at two levels: (1) coupling between individual subsystems (plant, cyber, actuation, sensors), and (2) coupling between nodes in a distributed computational subsystem. In this paper, a coupled smart system is decoupled and considered as a feed-forward system over time and modeled using a two-level Dynamic Bayesian Network (DBN), one at each level of coupling (between subsystems and between nodes). A DBN can aggregate uncertainty from multiple sources within a time step and across time steps. The DBN associated with a smart system can be learned using available system models, physics models and data. The proposed methodology is demonstrated for the design of a smart indoor heating system (identification of sensors and a wireless network) within cost constraints that enables room-by-room temperature control. We observe that sensor uncertainty has a higher impact on the performance of the heating system compared to the uncertainty in the wireless network.

Advances in data collection and storage infrastructure offer an unprecedented opportunity to integrate both data and emergency resources in a city into a dynamic learning system that can anticipate and rapidly respond to heterogeneous incidents. In this paper, we describe integration methods for spatio-temporal incident forecasting using previously collected vehicular accident data provided to us by the Nashville Fire Department. The literature provides several techniques that focus on analyzing features and predicting accidents for specific situations (specific intersections in a city, or certain segments of a freeway, for example), but these models break down when applied to a large, general area consisting of many road and intersection types and other factors like weather conditions. We use Similarity Based Agglomerative Clustering (SBAC) analysis to categorize incidents to account for these variables. Thereafter, we use survival analysis to learn the likelihood of incidents per cluster. The mapping of the clusters to the spatial locations is achieved using a Bayesian network. The prediction methods we have developed lay the foundation for future work on an optimal emergency vehicle allocation and dispatch system in Nashville.