Analytics in cyberspace uses the knowledge and useful information gained from data to feed optimal actions (or control schemes) back into the physical world. The Internet has evolved from wired computer networks to wireless, human-connected networks to the new era of smart and connected networks for making things. From the 1960s to the 1990s, the world saw rapid developments in Internet content materials such as emails, information, entertainment, web browsing, and HTML web pages.
Recently, we have witnessed the transition from the Internet of People to the Internet of Things. As the number of "things" connected to the Internet grows rapidly, the scalability of the protocol has proven to be a major challenge. There are many enabling technologies (e.g., cloud computing, virtual reality, IPv6, ambient intelligence) that contribute to the rapid development and implementation of IoT systems.
Sensor-based Manufacturing Informatics and Control
Based on the results of data analysis, the MES adjusts the manufacturing process (eg, predictive maintenance, operator shifts) to deliver work orders on time. Effective multi-sensor fusion strategies must consider both information transfer flows in real-time sensor signals and the evolution of nonlinear dynamics in the underlying processes. 62, 63] developed a new wavelet framework—multiscale recurrence analysis—to characterize and quantify the variations of nonlinear dynamics in the underlying processes.
Jing and Shi [66] proposed identifying causal relationships from observation data for manufacturing process control. Engineering knowledge was integrated with heuristic rules to learn arc directions in the causal network. Also, physics-driven models can be formulated based on specific failure mechanisms in the production system.
Data-driven models utilize the real-time sensor signals to characterize and model degradation behavior in the underlying process. However, discrete event simulation (DES) tends to track individual devices and their activities in the network of queues. As a result, DES models are not only time-consuming to execute, but also provide unrealistic approximations in the context of mass production or continuous manufacturing.
Furthermore, simulation optimization [74] can be integrated with the wealth of sensor data for production process modeling and decision support. The popularity score represents search interest relative to the highest point on the map for a given time in the world.
IoT Manufacturing Applications
- Case Study - IoT and Cloud Computing to Build Cyber-physical Manufacturing Networks
- A. Physical Machine Networks – Process Monitoring and Control
- B. Virtual Machine Networks
- C. Network Modeling and Analytics
- C.1 Pattern Matching
- C.2 Network Modeling
- C.3 Cloud Computing
- C.4 Predictive Analytics
- IoT and Cybersecurity in Manufacturing
- IoT Manufacturing Policies and Strategies
- IoT Challenges and Opportunities in Manufacturing
Safety and ergonomics: There are also many research efforts focusing on designing IoT systems for safety and ergonomics in the manufacturing industry. The second stage is feature extraction to characterize and quantify specific patterns in the IoMT data. However, the dimensionality of machine signatures is high and the number of machines is large in the IoMT context.
Therefore, we also present an idea of cloud computing for efficient network modeling of large-scale IoT machines in cyberspace, which will be detailed in the following subsections. As such, machines can form a community or group in the network that collectively provides a subnetwork of machines with similar characteristics. 6, IoMT connects a large number of machines in the production system and generates overwhelming big data.
Little has been done to address fundamental issues relevant to big data analytics in the broad context of IoMT. In the literature, such interconnections are assessed using methods such as correlation and mutual information. In the presence of a small number of machines (or profiles), the optimization of network node locations 𝒔𝑖 can be achieved with existing algorithms such as Multidimensional Scaling (MDS) [112] and Scaling with Complex Majorization Function (SMACOF) [ 113 ].
For example, we can locate each profile of a machine in the network clusters for monitoring or classification purposes. MTConnect advocates a read-only option when the top-level MES interacts with the smart manufacturing “things” in the IoT system [27, 28]. Please note that production data can be encrypted locally and in the cloud using the PGP standard [using entropy-generated keys and AES encryption] and transmitted via communications encrypted by hardware on-chip.
Smart manufacturing is also identified as an opportunity for Chinese manufacturers to take the lead in global competition.
Retrofit legacy machines for smart manufacturing
We observe that different machines can perform the same or different functions or tasks, and some machines rely heavily on the output of other machines, just like a pipeline product line. This status includes not only the fact of being busy or not, but also the state of health, in the sense of whether it is functioning properly or not. The simplest method is to use sensors that can both perform the sensing task and also provide some analysis based on signal processing of the sensed data.
However, with the increasing number of machines, given their life expectancy of one to two decades, it is difficult to provide wired power or battery support in some scenarios. Another challenge is how to take advantage of the status of different machines to distribute tasks to each machine. In these circumstances, dynamically distributing tasks to machines based on the perceived status of each machine is critical.
It should be noted that nowadays some tasks are performed by machines from different sources, even from different countries. This reveals the potential for communication reliability issues and its impact on collaboration between these devices. You can send G code to the machine to start it, but you cannot directly control the servo motors and machine spindles.
While old machines are an invaluable asset to manufacturing companies and are fully utilized in production, they lack real-time and in-process sensing and control systems. As a result, small producers are increasingly losing their competitive advantage in the global market because they are limited in information transparency and in their ability to manage the increasing complexity of modern production environments.
Self-powered machine status sensing
As a result, there is an urgent need to develop new plug-and-play IoT sensors that continuously collect in-situ machine data, transmit the data to cloud storage and communicate with other 'things' and stakeholders.
Machine service and tasks scheduling and distributing
The synergy between IoMT machines
Cloud computing and analytics
Blockchain enabled IoT
Conclusions
To achieve competitive advantages in the global market, the manufacturing industry strives to create new products and services. Note that while sensors, data, and IT systems may already be available in physical factories, they are not tightly integrated to the IoT level. Realizing the full potential of IoT for smart manufacturing requires new advances in analytical methodologies.
The challenges now are "how to mirror physical manufacturing in cyberspace through data-driven information processing and modeling?" and "how to exploit the useful information and knowledge extracted from data to provide better manufacturing operations in the physical world?". A number of IoT architectures such as RAMI 4.0 and OPC UA have been proposed to define the communication structure of Industry 4.0. The diverse types of IoT architectures and platforms are conducive to accelerating the development of IoT systems.
However, IoT is still under development and faces technical problems for cyber-physical integration in the manufacturing system such as communication, big data and control. For example, a single vibration sensor in the machine condition monitoring system generates data streams at high speed. Truth is also particularly important in the IoT paradigm given the uncertainty (and the lack of quantification of uncertainty) of statistical models.
Manufacturing researchers have traditionally been less concerned with the issues of big data analytics, cyber security, cloud computing, system optimization in the large-scale IoT context. This article provides an overview of the development of IoT technologies and existing applications in manufacturing enterprises.
Acknowledgements
Author Information
Fugee Tsung is Professor of the Department of Industrial Engineering and Decision Analysis (IEDA), Director of the Quality and Data Analysis Laboratory, at the Hong Kong University of Science and Technology (HKUST), and Editor-in-Chief of the Journal of Quality Technology (JQT) . He is a fellow of the Institute of Industrial and Systems Engineers (IISE), fellow of the American Society for Quality (ASQ), fellow of the American Statistical Association (ASA), academician of the International Academy for Quality (IAQ), and fellow of the Hong Kong Institution of Engineers (HKIE). He has written more than 100 journal publications and is also the winner of the Best Paper Award for the IIE Transactions.
Huang, “Big Data Analytics for Physical Internet-Based Intelligent Manufacturing Shop Floors,” International Journal of Production Research, vol. Perrone, "Business Process-Oriented Design of Enterprise Resource Planning (ERP) Systems for Small and Medium Enterprises," International Journal of Production Research, vol. Yang, “Heterogeneous repeatability monitoring of dynamic transitions in ultraprecision machining processes,” Journal of Manufacturing Systems , vol.
Hicks, "A Quantitative Method for Failure Mode and Effects Analysis," International Journal of Production Research, vol. Gunasekaran, “IoT-enabled servitization of manufacturing – an exploratory case study,” International Journal of Production Economics, in press, 2017. Zhou, “Opportunistic IoT: Exploring the harmonious interaction between human and the internet of things,” Journal of Network and Computer Applications, vol.
Spalanzani, "A bottom-up IoT-based approach to order fulfillment in a collaborative warehousing environment," International Journal of Production Economics, vol. Huang, “System dynamics analysis for an IoT-enabled manufacturing logistics system,” International Journal of Production Research , vol. Li, “Impact of RFID technology on supply chain decisions with inventory inaccuracies,” International Journal of Production Economics , vol.
Chang, "Ontwikkeling van prestasiemetingstelsel vir Internet van Dinge en slim fabrieksomgewing," International Journal of Production Research, vol. Kumara, "Parallel Computing and Network Analytics for Fast Internet-of-Things (IOT) Machine Information Processing and Condition Monitoring," Journal of Manufacturing Systems, vol.
