The definitions of cyber-physical systems (CPS) and the Internet of Things (IoT) are converging over time to include a common emphasis on hybrid systems of interacting digital, analog, physical, and human components in systems engineered for function through integrated physics and logic. CPS and IoT enable innovative applications and impact multiple economic sectors in the world-wide economy, including in energy infrastructures, advanced manufacturing, building control, transportation, health care, and others. Current design and management approaches for these systems are domain-specific, resulting in redundant efforts, limited sharing and interoperability across domains, and a lack of robust, formal methods for design, evaluation, verification, and validation. This project addresses these limitations through the development and application of a CPS Framework (applicable to CPS and IoT) to serve as a foundation for shared development, information exchange, and new formal methods applicable across domains. The CPS Framework provides an organized presentation of an analysis methodology based on the core concepts of facets (modes of the system engineering process: conceptualization, realization and assurance) and aspects (clusters of concerns: functional, business, human, trustworthiness, timing, data, composition, boundaries, and lifecycle) as well as integrated cyber-physical functional decomposition that together form a foundation for CPS and IoT.
Objective - To provide a common technical and conceptual foundation for CPS and IoT that enables conceptualization, realization and assurance across all domains, including a comprehensive and traceable methodology for meeting all stakeholder concerns throughout any systems engineering process and providing assurance that these concerns have been met.
What is the technical idea? Cyber Physical Systems (CPS), and increasingly, the Internet of Things (IoT), integrate computation, communication, sensing and actuation with physical systems and humans to fulfill functions of varying degrees of criticality and involving varying degrees of interaction with the environment. As described in NIST Special Publication 1900-202 titled “Cyber-Physical Systems and Internet of Things” (https://doi.org/10.6028/NIST.SP.1900-202), recognizing this convergence can bring currently isolated fields and sectors together for progress around shared research, application, and innovation goals and opportunities. Effectively designing, building, and assuring CPS/IoT systems requires consideration of the system’s functional context, including how the system is used and for what purpose or outcome.
A unified perspective on CPS/IoT systems allows a common classification structure for components, illuminating a path forward for enabling open composablity and reliable compositionality for innovation in the creation of novel systems and systems-of-systems applications. This unified perspective can be useful in prioritizing research, development, and deployment goals, including enabling tight physical and logical state linkages and developing hybrid discrete and continuous methods for conceptualization, realization, and assurance of CPS/IoT systems. The hybrid nature of CPS/IoT systems has important implications for engineering, including design assurance, cyber-physical security, lifecycle management, timing and synchronization, and more.
Collectively, these conclusions can inform research; commercial; standards; and legal, policy, and regulatory efforts designed to realize the value to society of advanced CPS/IoT technologies.
The basic elements of a CPS or IoT range from physical components, and their associated sensors and actuators, through control systems and analytics, to the overall optimization and user functionality, with assurance that requirements have been satisfied. The CPS Framework describes the work and work products of analysis, organized by groupings of activities or facets (modes of thinking: conceptualization, realization and assurance) and aspects (collections of concerns: functional, business, human, trustworthiness, timing, data, composition, boundaries, and lifecycle) that support system engineering analysis, design, development, operation, and validation and assurance of CPS/IoT. These analyses result in artifacts such as requirements, simulations of requirements, prototypes, project documentation, test plans and test results, standards and best practices and assurance cases.
Though most CPS/IoT today have a domain-specific focus, the future of CPS/IoT will undoubtedly include extensive communication and interoperation between domains to deliver coordinated functions. This CPS framework methodology provides a solid foundation for the development of a consensus technical architectures with common vocabulary and use cases that are, in turn, the basis for cooperation and collaboration for progress in all CPS/IoT applications.
What is the research plan? The research plan includes the following efforts: (1) progression of select components of the CPS Framework by NIST, such as in the areas of trustworthiness and assurance mathematics and formal methods; (2) application of the CPS Framework to selected vertical domains, including the transportation domain in collaboration with industry; (3) development of models and tools that precisely formalize the functionality of the CPS Framework, with use case modeling to support technology transfer to industry, including the transportation industry; extension of these models to enable aspect/concern-based dashboards for use cases to assess current status and mission competence; (4) extension of the ontology models for design- or operation-focused reasoning to assess whether stakeholder concerns are met and, if not, what action can be taken to maintain mission competence while recovering complete or partial compliance to stakeholder concerns; and (5) continued introduction of technical contributions to and engagement with relevant national and international research and standards efforts.
Following the publication of the CPS Framework, the project is focusing on more targeted work to improve key components of the Framework, such as trustworthiness (safety, security, privacy, resilience and reliability), CPS Framework modeling as an ontology, ontology- based methods for the analysis of dependencies between aspects (such as formalized reasoning systems), and formal methods for assurance. As part of this activity, additional expertise in areas such as safety and reliability will be identified to further develop the aspect of trustworthiness of the CPS Framework, with the goal of developing and publishing an expanded CPS Framework component (aspect) of Trustworthiness.
Through collaboration with industry, an application of the CPS Framework to the transportation domain is being produced, with special attention to trustworthiness for autonomous/connected vehicles. In addition, an application of the CPS Testbed co-simulation platform, UCEF (Universal CPS Environment for Federation), will be demonstrated for the transportation domain, as described in the CPS/IoT Testbed project. The UCEF-based autonomous vehicle CPS Testbed will enable the study, using CPS Framework modeling, of autonomous vehicle control functions, including artificial intelligence algorithms, against existing models of vehicle dynamics and the assessment of these functions relative to CPS Framework trustworthiness concerns, including safety, security, privacy, resilience and reliability. Interdependencies between these concerns have been studied using the ontology/reasoning model of the CPS Framework and analysis of UCEF-based experimental data may reveal others. Based on this work, a publication is planned to summarize the use of UCEF technology for transportation systems and developing the new measurement methodologies needed to assess the trustworthiness of future autonomous systems.
The CPS Framework modeling activity has produced UML, XML and ontology models that enable mathematical, data-exchange and formal reasoning respectively. The UML model in turn was composed with the UML use case model (IEC 62599-3) and was used to generate an XML model, comprising a CPS Framework-based data structure overlaid on the use case data structure. Using this XML model, an example was constructed of the conceptualization artifacts. In this way, a versatile data structure can be defined that is aligned with the CPS Framework and can be used for data exchange and tool integration during the systems engineering process of designing, building and testing a CPS. Outreach to system engineering tool builders will continue in FY20 to encourage uptake of the CPS Framework and its technology by industry through incorporation of these tools into their own development tool suites.