Cyber-Physical Systems (CPS) combine the cyber and physical worlds with technologies that can respond in real time to their environments. CPS and related systems (including the Internet of Things, Industrial Internet, and more) include co-engineered interacting networks of physical and computational components - examples include a smart grid for clean, efficient and reliable energy; intelligent, wearable medical devices for better health and an improved quality of life; autonomous vehicles that increase safety, decrease congestion, and reduce transportation costs; and interacting CPS systems, such as smart emergency response working cooperatively with smart traffic networks to control traffic flows and enable faster transit of emergency vehicles to incident sites and medical facilities. The CPS Program develops and demonstrates new measurement science and promotes the emergence of consensus standards and protocols for advanced systems that are reliable, resilient, effective, safe, sustainable, secure, and privacy enhancing. NIST-wide program coordination is provided by the Engineering Laboratory (Smart Grid and Cyber-Physical Systems Program Office) and the program also draws on the expertise of the Information Technology and Physical Measurement Laboratories.
What is the problem?
Measurement science is lacking to support design, development and deployment of composable, scalable, and interconnected CPS systems in and across multiple "smart" domains, including in complex smart cities environments. The President's Council of Advisors on Science and Technology has identified cyber-physical systems as a national priority for federal R&D. Deployment of next-generation CPS across the transportation, energy and health sectors alone could boost U.S. productivity growth by as much as 1.5 percent, according to some estimates. The implementation of new cyber-physical systems to achieve just a one percent improvement in efficiency can save $30 billion in aviation sector fuel costs, $66 billion in power generation, $63 billion in health care and $27 billion in freight rail costs over a 15-year period. To realize the full economic benefits of next-generation CPS, multiple challenges must be addressed. The design and engineering of a cyber-physical system, from initial concept through successful operation, requires a new systems science and engineering approach. This approach must simultaneously embrace all levels of the CPS architecture, from physical components and their associated sensors and actuators, through control systems and analytics, to the overall optimization and user functionality. Advanced cyber-physical systems are so complex that existing approaches for performance prediction, measurement and management do not apply. And much current CPS work is done in isolation, focused on solutions limited to a single domain such as health care or manufacturing, with limited cooperation across the commercial, academic and government sectors.
What is the technical idea?
The key technical ideas can be summarized as follows.
- The first measurement science problem is the need for a credible technical architecture suitable to the full range of CPS use cases. The research plan provides for the development of a CPS reference architecture that enables collaboration among stakeholders, discovery of common principles applicable to many CPS implementations, and the identification of critical gaps in standards and metrics.
- The second measurement problem is the need to integrate work and share ideas and solutions across a broad range of disciplines and domains. Development of the CPS Framework and its analysis methodology through a public consensus process creates a sense of community, shared purpose, and teamwork at a level that will be needed to take on the complex challenges inherent in cyber-physical systems.
- The third measurement science problem is the need for a platform for CPS experimentation and validation. The research plan provides for the development of a modular CPS testbed to support NIST measurement science development for CPS. Integration of composability and modularity in the design of the CPS testbed allows its application to evaluating performance of CPS systems in multiple domains, enabling its use by a diversity of communities for a range of applications, and demonstrating its agility and application at large scale through reconfigurable combinations with other testbeds at NIST, across the nation, and around the world.
- The fourth measurement science problem is the need for NIST to lead in organizing governments/users (cities) and technology innovators (industry and academia) to demonstrate a scalable and reproducible model for incubation and deployment of interoperable, adaptable and configurable IoT/CPS technologies and solutions in Smart Communities/Cities. The use of Challenge initiatives, such as SmartAmerica and Global Cities Teams Challenge, further creates a sense of community and shared purpose, creates new teams of innovators and adopters, mobilizes academic, commercial, and government resources toward shared objectives, facilitates the identification of standards and measurement needs, and highlights NIST's role as a neutral convener and technical expert in the CPS field.
What is the research plan?
The research plan comprises two elements as follows. The first focuses on new approaches enabling the design and engineering of a cyber-physical system from initial concept through successful operation. This requires a new systems science and engineering approach. This approach must simultaneously embrace all levels of the CPS architecture, from physical components and their associated sensors and actuators at the base layers, through middle-layer control systems and analytics, to the overall optimization and user functionality at higher layers. Principles for integrating these layers in a scalable design strategy include:
- integrating concepts from engineering, information technology, physics, and materials science; ″providing for interoperability, modularity, and composability;
- enabling designed-in trustworthiness in resilience, safety, cyber- and physical-security, and privacy;
- interconnecting data and analytics across levels and between systems through networking; and
- supporting all phases of the design cycle from initial concept to manufacturing and deployment.
The research plan applies these principles in two key areas to enable new, scalable CPS design approaches. The first area is the development of a common vocabulary that enable shared progress across current, siloed CPS domains. These include a reference architecture, syntax, and ontologies that provide the basis for modeling, programming, control, and communications languages that span domains and disciplines. This work provides the essential foundation for subsequent development of standards for interoperability and composability across architectural layers and between components and systems. The second area focuses on security and privacy status during operations and includes consensus guidelines and measurement processes for security automation and quantitation, privacy, and high-confidence networks with assured quality of service. The results are essential to developing CPS for use in sensitive applications such as health care and assisted living; in safety-critical applications such as remote surgery; in time-critical applications such as the smart grid; and in critical infrastructures for disaster resilience, traffic management, and municipal water systems.
The second element of the research plan focuses on the capabilities required for experimental manipulation, measurement and evaluation of the performance of the more capable and powerful cyber physical systems enabled by the new design approaches targeted under the first area. In this context, CPS performance metrics include efficiency and sustainability, agility and flexibility, reliability (including time critical performance), resilience, usability, safety, security, and privacy. Research in this second area focuses on the development of a comprehensive abstraction infrastructure comprising tools, platforms, testbeds, and integrated design environments to enable the application of formal methods and standards to the co-design of heterogeneous, interacting components. Testbeds and research platforms developed under this initiative will be modular, reconfigurable, remotely accessible, and adaptable to multiple domains and applications.