The design of the new CPS/IoT System will depend on the application of various computational models and protocols, including e-work parallelism, Task Parallelism Optimization Module (DPIEM-2), Keep it Simple (KISS), Emergent Lines of Collaboration and Command (ELOCC), and Best Matching (BM) principle. These methodologies are part of the collaborative control theory (CCT) and their continued utilization to address diverse engineering and design challenges. The realization that the intrinsic sustainability, opportunities, and advantages of the emerging e-activities and E-Systems will not materialize without an effective design through collaborative engineering underpins this work.
E-Work Parallelism (EWP)
The e-work parallelism system exploits the fact that in any distributed e-system network of agents, some activities can and should be performed in parallel. This principle has deep ramifications due to the agent’s distributed nature and the interaction between software and human workspaces (Nof et al., 2015). It covers computer-computer, human-computer, and human-human interactions. EWP is performed and planned using the Distributed Parallel Integration Evaluation Model (DPIEM). This approach is necessary for task-based integration workflow optimization. The approach considers operational constraints, task dependence, and information availability. DPIEM has three modules (DPIEM-1, DPIEM-2, and DPIEM-3).
Task Parallelism Optimization Module Two (DPIEM-2)
This module considers the economic aspects of e-work parallelism (EWP) by optimizing the integrated agent’s total cost. The two main components of this total cost are communication and operational costs. Communication costs cover all non-operational costs that cover task execution at each node (Nof et al., 2015). The three components of the communication costs are sending the information by the source task, network delay, and receiving the information by the target task. Operational cost refers to costs that cover task execution at each node. This computational and design approach limits technological resource utilization, making systems more effective and robust.
Keep It Simple, System (KISS)
E-system complication is not always desirable as this can be costly and time-consuming to implement. However, systems are allowed to be as complicated as they need to be, provided they are as simple as possible. Keeping systems as simple as possible is necessary to facilitate effective and useful interactions with human participants. The KISS principle builds on and embraces traditional human automation and human-computer usability design principles (Nof et al., 2015). The KISS system to be used in the proposed CPS/IoT system is continuous improvement, which allows for the ongoing eradication of complicated components. The system will also be an evolutionary learning one, which comprises internal machine learning mechanisms that facilitate pattern recognition and information analysis.
Emergent Lines of Collaboration and Command (ELOCC)
The main goal of the ELOCC system or design approach is to overcome the drastic changes that networks may face in emergencies and the volatilities of informal and formal communications between clustered agents and individuals. This approach ensures that the system is more resilient and reliable to increase user experience and satisfaction. The approach also lowers the operational and maintenance costs because distributions would not always result in significant system destruction or disturbances (Nof et al., 2015). The ELOCC approach will ensure the developed system finds adjustment when the network evolves as expected. It facilitates gradual changes and more effective decisions, sustainable information exchange, knowledge creation, and emergent network control and optimization. It is the needed approach for continuous improvement within the uncertain and dynamic information technology environment.
Best Matching (BM) Principle
The best-matching principle is a critical consideration in designing and executing collaborative e-work networks. This approach presents a challenge to the systems designer by matching agents within a network to facilitate effective collaboration and interaction. Ensuring that agents are connected to the best match increases operational efficiency and reduces the associated cost (utilization of information technology resources) (Nof et al., 2015). This principle will require the comprehension of the characteristics of each agent and the possible impact and ramifications of various connection possibilities. The simplest case is analogous to the original one-to-one assignment problem.
System Problems to Resolve
Cyber-physical systems (CPS) and internet-of-things (IoT) technologies have significant applications in various areas and domains. These days, engineers use them more frequently to design smart cities, smart buildings, smart homes, power distribution grids, transportation systems, and intelligent greenhouses (Bu et al., 2018). Well-designed CPS/IoT increases building comfortability, enhances safety, and reduces energy consumption. Smart buildings and cities are use-case scenarios for CPS/IoT, but problems exist in developing and implementing effective and reliable systems. These concerns are in five critical areas: virtual run-time environments, data quality, fault tolerance, computation model, and middleware. Designing an effective and highly operational CPS/IoT system will depend on resolving these problems.
Middleware
Middleware is a useful technological component and is necessary for service development and management. Many CPS/IoT service providers eliminate the burden of service management by seeing end devices as dumb and with little or no computational abilities (Tantawy et al., 2020). They deploy most, if not all, of the services to the internet servers. A communication gateway transmits the collected data to the servers. These communication gateways also facilitate the downloading of the control commands. Designing better middleware eases the management of CPS/IoT applications and eradicates user constraints. The middleware’s built-in intelligence performs sensor detection, system reconfiguration, software deployment, system configuration, and device selection.
Data Quality
The second challenge that smart cities and building CPS/IoT computing services experience concerns data quality. In physical systems, the validity of data depends on the sensed time. It is also difficult or impossible to store this sensed data unless there is a provision to allow it. As such, there is a need to properly present the data, paying attention to its quality and including information about the sample time, location, lifetime, and accuracy to ensure that the computing process does not misuse or mistreat physical process data (Nof et al., 2015). After confirming the credibility and reliability of the data, it will be necessary to ensure that the CPS/IoT system confirms that the available data meets the computing process system’s requirements. The data must always be accurate, reliable, fresh, and correct.
Service Composition and Computing Models
The aim and objective of deploying sensors in smart homes and cities are data collection and responding proactively, actively, or passively depending on what the information reveals. In this regard, the fundamental question when designing a smart city and building CPS/IoT is how to process the collected data. Possible data processing options include centralized methodologies, distributed approaches, on-edge devices, and on-cloud devices (Macchi, 2019). For example, SmartThings uses a centralized approach to data processing. The problem with the centralized approach is that it is less scalable and less responsive. A distributed data processing approach can resolve these problems. However, distributed data process is less straightforward and more complicated; the devices may not even require internet connectivity. Data transmission, delay tolerance, and algorithm design issues impact distributed systems.
Fault Tolerance
The fourth challenge for smart cities and building CPS/IoT is fault tolerance. It is necessary to resolve this problem to ensure a more reliable and effective system that achieves its goals (Kang et al., 2019). The fault-tolerance problem happens because IoT device failures may disrupt or endanger a person’s life. The heterogeneity in computation and communication capability makes it not trivial to detect faults.
References
Bu, L., Zhang, T., Chen, X., Wang, L., Zhao, J., & Li, X. (2018). Model-based construction and verification of cyber-physical systems.ACM SIGSOFT Software Engineering Notes, 43(3), 6-10.
Kang, B. G., Seo, K. M., & Kim, T. G. (2019). Model-based design of defense cyber-physical systems to analyze mission effectiveness and network performance. IEEE Access, 7, 42063-42080.
Macchi, M. (2019). Revolutionizing collaboration through e-Work, e-Business, and e-Service. Taylor and Francis.
Nof, S. Y., Ceroni, J., Jeong, W., & Moghaddam, M. (2015). Revolutionizing Collaboration through e-Work, e-Business, and e-Service (Vol. 2). Springer.
Tantawy, A., Abdelwahed, S., Erradi, A., & Shaban, K. (2020). Model-based risk assessment for cyber physical systems security. Computers & Security, 96, 1-15.