1.3.1 Defining Cyber-Physical Systems
In a cyber-physical system (CPS), some elements are computational and some involve interactions with the physical environment [8, 9, 10, 11, 12 and 13], integrating ācomputation, communication, sensing, and actuation with physical systems to fulfill time-sensitive functions with varying degrees of interaction with the environment, including human interactionā [1,14]. A CPS incorporates multiple connected systems, producing a system capable of developing an awareness of its physical environment and context [15], making decisions based on that information, and enacting work that can effect changes in its physical environment [16].
As an example, consider a traffic management system (TMS). In many jurisdictions, road networks are divided into regions, each controlled by a separate autonomous TMS. The TMS is intended to meet several goals, some of which may conflict. These can include, for example, ensuring optimal throughput with minimum congestion, improving road safety, reducing air pollution and fuel burned, ensuring consistent travel times, etc. The TMS relies on data transmitted by large numbers of monitoring devices that are typically installed roadside or buried under the road surface and connected to a local traffic control center (TCC). The TCC conducts analysis, making predictions based on current data about likely congestion in the near future, identifying current problems or hazards, and suggesting appropriate strategies. Decisions made by the TCC are communicated to a variety of further roadside devices that can influence traffic behavior, such as variable speed limits and message boards, dynamic lane closures, and variable timings on traffic lights.
This is an example of a large-scale CPS; it relies on devices that can observe or affect the real world, gathering data from sensors, analyzing it, and making adjustments as necessary to improve performance. It enables a flexible solution that identifies problems and quickly adapts (e.g., by imposing speed limits or opening extra lanes). However, it is a complex system with an enormous variety of sensor types (and therefore significant heterogeneity), as well as complex analysis and data visualization. The application domain demands a high degree of dependability, which in turn is reliant on the behavior of different participating systems, from sensors to communications systems to analysis algorithms. Dependability includes real-time requirements; the situation on the road can change relatively quickly, and if analysis takes too long, the recommendations produced will be based on out-of-date information.
This traffic management example provides an illustration of a CPS in one domain, but the same principle of combining sensors, actuators, and intelligent analysis can be used to build CPSs that deliver improved performance, flexibility, or efficiency in many other domains. For example, assisted living systems can rely on wearable sensors or nonintrusive devices installed around a building to identify when an elderly person who lives alone needs help. CPSs can be used in manufacturing to monitor quality and make adjustments automatically that improve performance and reduce waste or allow a manufacturing line or other industrial process to adapt dynamically to volatile requirements. CPSs are suitable for these domains and a wide range of others.
CPSs can cross organizational boundaries, with one or more organizations contributing constituent parts toward the whole. In addition, a CPS crosses multiple engineering, computer science, and social science disciplines by incorporating elements that interact with the real world, human systems, and complex software systems capable of intelligently processing the large amounts of data that CPSs may encounter [9,17].
The TAMS4CPS definitional framework [5] points out a variety of definitions that exist for CPSs. For example, some define CPS as āintegrations of computation and physical processesā [18] or āsmart systems that encompass computational (i.e., hardware and software) and physical components, seamlessly integrated and closely interacting to sense the changing state of the real worldā [19]. Other definitions emphasize the ācyberā aspects of CPS engineering, for example, defining CPS as
āICT systems (sensing, actuating, computing, communication, etc.) embedded in physical objects, interconnected (including through the Internet) and providing citizens and businesses with a wide range of innovative applications and servicesā [20].
āComputation, communication and control components tightly combined with physical processes of different nature, e.g., mechanical, electrical, and chemical. Typically a CPS is defined and understood (evaluated) in a socia...
