Research Trends and Performance of IIoT Communication Network-Architectural Layers of Petrochemical Industry 4.0 for Coping with Circular Economy

,


Introduction
In the last decade, the demand from customers for customized products with the lowest prices thrust Production Systems (PSs) into more digitalized capabilities with an effort of minimizing the wastefulness of industries. This approach is called the Circular Economy (CE). The growth of the population also propelled industrialists to establish, expand, and integrate their business empires by incorporating different varieties of customized products with the least waste. This is another way to describe CE. Today's emergent demands of domestic oils and gas resources are stimulating the research communities towards devising appropriate means to monitor and evaluate the physical performance for saving energy, processes which are respected as CE as per [1,2]. Firms adapt a technical and institutional framework for enrolling, recycling, and disposal monitoring systems to ensure utilization of waste oils and hazardous wastes with a huge public interest in developing CE [3,4]. The employment of appropriate technology can radically reduce ecological pollution and wastes by improving connectivity, integration, and automation towards oil mill processors and allied systems for attaining CE [5,6].
Today, it is necessary to react towards technological advancements and deployments for coping with the demand for liquid fuel, the shortage of crude oil, monitoring of liquid fuel routes, and the assessment of their life cycle's estimation for addressing CE as per [7,8]. Industries ought to focus on stimulating the biomass means, technical routing, byproduct synergy, carbon capture features, system automation, scientific implications, feedback mechanisms, connecting nodes, etc. in industries for attaining global CE as per [9][10][11]. Additionally, the diverse varieties of emissions of offensive gases from production should be managed and monitored by developing national strategies, control systems, technology, etc., as they can cause irreparable loss, wastes, and commensurate costs to the society and to industrial plants [12,13]. Accordingly, technological proximity and means need to evolve and be incorporated into a firm's system by modeling and integrating to build CE as per [14,15]. An elevated level of connectivity, integration, and monitoring is evidently needed among business networks and is required for coping with the global CE, competition, and management of production costs as per [16,17]. High-rank digitalization mechanisms, connecting technologies, and automation are needed at every corner of the organizational structure for attaining CE as per [18][19][20].
It is ascertained that the industry 4.0 revolution captured entire industries and markets around the world. Industry 4.0 is described in terms of automations, information exchangers, cloud storages and computing, cyber tech and physical systems, robotic self-navigation phenomena, big data analysis, and IoTs to add values in flexible and efficient PSs. Technological advancement, innovation, evolution, etc., are required for effective interconnection and communication among different manufacturing systems, which are demanding the critical exploration of industry 4.0 as per [21,22]. The reactions of digital CNs/ALs are propelling the significant advancements in industry 4.0 structures and are assisting industry 4.0 for driving PSs with flexibility with least waste output. Industry 4.0 introduces electronic data feedback and an integrated digital medium system across the functional units of industry. Industry 4.0 stimulates the computerization, monitoring, automation, and setting up of wide CNs for predicting and self-diagnosing deployed manufacturing devices. Ennis et al. [23] reveals that the collaboration and integration among organizations are possible today due to industry 4.0. Oztemel and Gursev [24] stated that industry 4.0 can make PSs more capable for performing activities and achieving CE. Yuan et al. [25] suggested that industry 4.0's practices are the best for smart manufacturing and advised that industry 4.0 is a reliable and scalable platform (digitally operating hardware and software) to update the level of instrument technology for effective utilization of resources. Trotta and Garengo [26] and Maresova et al. [27] found that industry 4.0 is a complex and disruptive network, which helps the stakeholders of private and state companies to attain more revenue and develop the global CE.
It has been empirically investigated that Petrochemical Industries (PIs) have been counted as one of the significant sectors that fulfill the daily needs of society's people. Presently, PIs are leading and immensely expanding their business globally by producing chemical compounds and petroleum-based solvents such as paints, coatings, detergents, and adhesives. Nowadays, most of the PIs are emergently focusing on manufacturing beverage products, i.e., oil and gas refinery, vegetable and palm oil, petroleum oil, ethylene refinery, and ink and paint production for attaining a strong CE. [28][29][30]. A few PIs turn out ethane, ethylene, methane, and petrol to compensate for the daily needs of peoples. Today, PIs are engrossed into the industry 4.0 revolution or digitalization, which is called PI 4.0 . The majority of PI 4.0 becomes sustainable and more lenient to flexible, agile, and lean manufacturing processes on account of digital transformation for attaining a strong CE. PI 4.0 overcomes the prohibitive issues related to production and monitoring of PSs with great digital network and capabilities, which help industries meet the challenges of market demands.
Investigations have shown that the Communication Networks/Architectural Layers (CNs/ALs) with PI 4.0 (which is abbreviated as PI 4.0 -CNs/ALs), such as IoT, CPS, VR, I, and DO, build the PSs of PIs more lenient to CE and the market's demands. Therefore, the continuous digital advancement into CNs/ALs enabled PI 4.0 to meet the future market's demands. However, there is a need to map the RTs against CNs/ALs of PI 4.0 to identify the areas where CNs/ALs are still poor, weak, and not at all improved. This rationale diverts the attentions of the authors to quantify the RTs of PI 4.0 -CNs/ALs. Gölzer et al. [31] argue that RT materialization is needed for a complete and detailed description of data processing requirements for all industry 4.0-based companies, to fully understand implications on IT-infrastructure and IT-solution-components. Van Tienen et al. [32] found that the digital advancement solution against CNs/ALs is a simplistic approach to improve industry 4.0 technologies and entrepreneurships and attain a high degree of CE. Groger [33] suggested that RT directions are useful for the successful implementation of industry 4.0, i.e., interdisciplinary research, specification of modular and reusable analytical services, appropriate tools, and organizational models and frameworks.

Literature Survey
The CNs/ALs augmented the convenience, handiness, and productivity across PIs due to intelligent interpretation of procured information and consequently aided the autonomous process [34,35]. The apparent requirement for advanced digital technologies belonging to Internet of Things, cloud computing, information exchangers, etc., is embodying a novel paradigm and influencing countless aspects of the routine line of private and business users [36,37]. Theoretical supports and academic researches provided the comprehensive pitch to be explored by smart industry 4.0 [38][39][40]]. An assessment model was built in [41], and it was used to measure and gauge the industry 4.0 technologies by considering three dimensions such as factory of the future, people and culture, and strategy. The eight chief attributes, i.e., cloud computing, manufacturing execution system, evalue chains, additive manufacturing, IoTs and cyber physical systems, big data, and sensors and autonomous robots are suggested as "Factory of the Future" and implemented 2 Wireless Communications and Mobile Computing as the significant dimensions of industry 4.0. Accordingly, enablers of industry 4.0, interconnected nodes and medium, research questions, technical frameworks, distribution patterns, computation mechanisms etc., are respected as effectual patterns of industry 4.0 evolution [42][43][44]. A progressive elementary aspect can assist in visualizing, monitoring, and recovering the characteristics of a manufacturing system [45,46]. The revolutionary concepts and creative tools are required to identify the significant enablers for future success of organizations [47,48]. Multiple flow control mechanisms and flexibility in production, decision-making, and problem solving are needed for executing the production processes efficiently based on the operational models [49]; consequently, a novel operational model called the Internet Fulfillment Warehouses (IFWs) was devised in [50] for effectively optimizing the operational model of warehouses.
Internet is the key for technical innovation, and it possesses the capability to synchronize and optimize the static and dynamic constraints to enable the technical execution of industry 4.0 [51,52]. Internet is presently serving as a novel operational configuration by linking smart apparatuses, machineries, and systems [53,54]. Comprehensive integrated tools, models, strategies, policies, and techniques need to be integrated with industry 4.0 when tackling human resources [55,56].
The authors adapted the internet-based research search engine [57] to accumulate databases related to research documents. The internet-based search engine has evidently been adapted in previous research works and has eliminated the drawback of the expert's panel and furcating schemes. The authors further conducted a Systematic Literature Survey (SLS) to recognize the momentous CNs/ALs of PIs and to quantify the RTs of PI 4.0 -CNs/ALs. Matthew et al. [58] emphasized the failure detection analytics sensors to monitor the oil production failures in digitized oil fields. The research work suggested the use of thousands of sensors and gauges with equipment to map the physical and chemical characteristics of oil and gas extracted from underground reservoirs. John and David [59] proposed a simulation model to fruitfully analyze the dynamic magnitude of water and sand collected by extracting the oil from a pool of oil. An analysis was conducted to choose the optimum technology for coping with oil extraction challenges. Li et al. [60] proposed extensible X-3D software for building an interactive and dynamic virtual oil and gas pipeline system. The X-3D application was applied in designing a piping system to illustrate the virtual reality application in PI 4.0 . Meng et al. [61] stated that IoT is a significant indicator for taking care of information technology and is respected as a crucial AL for executing industrial operations effectively. The authors demonstrated an IoT AL reference model to investigate IoT growth. A case study of a PI 4 . 0 is illustrated to reveal the research challenges and opportunities associated with IoT.
Hemant et al. [62] proposed an infeasibility driven evolutionary algorithm to resolve the efficiency evaluation problem of fifty-six oil pools under a single-level constraint variable. The authors have suggested to frame a multiobjective problem to define effective results and to eliminate the drawback of an earlier formulated objective. David [63] explained a key CNs/ALs of PI 4 . 0 proposed by the American National Standards Institute, the American Petroleum Institute, and Standard 780 to build security risk assessment methods for systematically identifying suitable measures and eliminating future threats. Parolini et al. [64] proposed a new Cyber Physical Index (CPI) for measuring the effects of a combined distribution of a Cyber Physical System (CPS) in a given data. A case study is conducted by the authors to demonstrate how the CPI indicates the potential impact of CPS control strategies and cyber cum physical control as well. Gholian [65] developed a mathematical model for establishing the optimized operational sequences for industrial load control operation. Yatin and Clifford [66] proposed a game theory algorithm to allocate the cybersecurity controls in the oil pipelines. The proposed algorithm assisted the oil pipeline cyber physical system to allocate the cybersecurity control teams around high-risk regions. Ahmed and Kim [67] described the Named Data Networking (NDN) with its applications in smart home communications for critically evaluating and defining the aspects to address the future challenges of NDN.
Robin and Chunyan [68] utilized the ERP (Enterprise Resource Planning) system and suggested its implementation over Continuous Auditing (CA) of oil refinery processes. It has been concluded that ERP increases the efficiency, fraud risk reduction, knowledge application, and credibility of the auditing team. Jeon et al. [69] proposed a specific plan to effectively implement ERP for controlling the shop floor of PI 4.0 . In the proposed plan, an advanced MES is added for collecting, measuring, and analyzing shop floor controlling. Niggermann et al. [70] determined that data-driven approaches to analysis and diagnosis of CPSs are always inferior when compared with classical modelbased approaches, constituted by experts. Trappey et al. [71] evaluated the critical international standards and intellectual property rights (associated with CPS patents) to benefit academic scholars and industry practitioners. Hassani et al. [72] focused on the evaluation of the significant innovations, technological drivers, and CNs/ALs of PI 4.0s . The authors searched the quantifiable and nonquantifiable impacts of innovation, technological drivers, and CNs/ALs that benefit PI 4.0s .
After passing through the aforesaid literature survey, the authors built a research methodology and listed four Research Inquiries (RIs) to effectively grab the significance for commencing the research work. The four research inquiries and manuscript filtering/screening guidelines for quantifying the RTs against PI 4.0 -CNs/ALs are framed. The authors reported a research approach and structure for successfully materializing the RTs and suggesting the future research challenges of PI 4.0 -CNs/ALs to attain CE. repairing, refurbishing, remanufacturing, and recycling systems in production systems of companies). CE develops economic, natural, and social capital by addressing the three challenges of companies, i.e., concentrating on elimination of waste and pollution, recycling products and materials, and generating energy. In the present PI 4.0 , CNs/ALs contribute towards balancing the advance of manufacturing and the reutilization, recovery, remanufacturing, and recycling (reverse manufacturing) of scraps/parts aiming at forming the CE of industries. CE improvement across PI 4.0 can be attained by enriching and advancing the CNs/ALs through identifying the RT levels. In the research forum presented in this study, the authors built and proposed a research method and four Research Inquiries (RIs) for quantifying the RTs of PI 4.0 -CNs/ALs. The research structure is illustrated in Figure 1

CPS
It integrates computational algorithms with physical elements, i.e., robot and CNC of production system. CPS is a consistent synergy between hardware and software towards driving the efficient production systems.

VR
It demonstrates the practical experiments before commencing work in reality. VR passes the investigators through a phenomenon before conducting the same phenomenon in reality. ANSIS, CAD, POE, VERSA CAD, KEY-CAD, and AUTO-CAD assist the investigators/designers to realize the production before commencing the same in reality/practically.

I
It is a technique to increase the future business opportunities. It focuses on augmentation of the firm's capacity at the market, eliminates the transaction costs, and secures multiple vendors or distribution channels for the future.

DO
It deals with evaluation of the most excellent choice under cost and other business parameters. In case of DO, the business parameters are interpreted by simulations, models, decision support systems, and decision-making tools.

ERP
It is a computerized system, focuses on controlling the business operations such as customer orders and tracking, scheduling operations, reviews inventory records, and prepares the financial sheet related to production and marketing.

PC
It deals with storing and retrieving the input and output data via electronic signals for maintaining the production assets for future use.

DA
It deals with interpretation of the input and output digital data by exploring the vector machine algorithm, clustering algorithm, support, linear regression, logistic regression, artificial neural networks, and sensors to enable the production system to work without failures.
N It interconnects the soft computing and peripheral devices with parent's identical devices. The cross-functional units of production systems are connected by networking.
IDM It helps for simulating the data/information for planning, modeling, security control, conducting experiments, data analytics, and quality control purposes. It escalates the data sharing and fosters the information to other departments for use.    [57], which are targeted in this research to store the relevant database and thus satisfying RQ 2 . The authors collected only DOIs/URLs against PI 4.0 -CNs/ALs over years 2007-2017. The authors explored primary, secondary, and tertiary protocols as discussed in Table 6 for searching research documents.
The authors applied the Sum of Digit (SD) technique to quantify the research manuscripts, published over the years 2007-2017. Table 7 summarizes the results against the scientific research databases I and II with respect to the exclusion parameters, full text search, and primary search. Figure 2 illustrates the database using the Sankey flow diagram, which shows the total research documents and its scattering record. Figure 3 reveals the total database of PI 4.0 -CNs/ALs by bar chart. Figure 4 evaluates the research documents by PRISMA 2009 flow chart, which shows the research documents' refinery process (inclusion and exclusion documents) and the research documents to be considered for studied, quantitative, and qualitative analysis.

Result and Discussions
The line charts are presented in  Paint production/industry Ink production/industry Glycol production/industry Oil/petroleum refinery production/industry Tertiary search on oil production under aspects Natural gases Coal Waste products Agriculture waste  After identifying as well as discussing the RTs of PI 4.0 -CNs/ALs, the authors focused on suggestions, provided by published articles in improving the very weak and moderately performing PI 4.0 -CNs/ALs, linked to CE. The authors present Section 5.1, which directs the scholars towards conducting research over very weak performing PI 4.0 -CNs/ALs such as CPS-ERP-IDM and subsequently focusing on the moderately weak performing PI 4.0 -CNs/ALs such as N, DA, PC, IoT, VR, and I.

Suggested Research Areas to
Be Focused for Improving the Very Weak Performing PI 4.0 -CNs/ALs Such as CPS-ERP-IDM. Asongu and le Roux [73] and Miksa et al. [74] focused on the information and communication system and advised scholars to focus on information technology to enable effective communication systems in IT sectors. The presented research suggests that future scholars should focus on electronic devices, security risks, wireless monitoring control, knowledge and big data management, maintenance systems, and smart manufacturing architectures for improving the future performance of PI 4.0 -IDM-CN/AL and CE.
Robin and Chunyan [68] investigated the ERP system of oil and gas industries in Houston and analyzed the RTs and

Wireless Communications and Mobile Computing
RTs of models, mathematical modeling, and applications of algorithmic techniques under I 4 structures. The authors advised scholars to focus more towards path planning, machine learning process, and ERP software benefits for enhancing the future performance of PI 4.0 -ERP-CN/AL. Mbohwa and Sahu [75] suggested that researchers should work on cyber physical security risk, cybersensor nodes, application of CPS principles, and polymorphic wireless receivers to improve the future performance of PI 4.0 -CPS-CN/AL and CE.

Suggested Research Areas to
Be Focused for Improving the Moderately Weak Performing PI 4.0 -CNs/ALs Such as N, DA, PC, IoT, and VR, I. Lu [42] suggested that scholars should focus on such areas as development of models and data modeling, application of techniques/methods/algorithms, fuel market integration, integration of biofuel filtration, new technology, the best decision styles, and design integration and vertical integration in order to improve the future performance of PI 4.0 -I-CN/AL and CE.   Wireless Communications and Mobile Computing Nazari et al. [76] advised researchers to follow up areas such as technological aspects, virtual reality architectures, analytical simulation or virtual testing of oil dynamic aspects and application of software model development, fault monitoring and diagnostics, Java-based toolkit, proportions, virtual line process monitoring, virtual reality-based education program, analytical simulation for parameter optimization, and data sharing to social network websites for escalating the future performance of PI 4.0 -VR-CN/AL and CE.
Meng et al. [61] and Celia and Cungang [77] proposed that scholars should focus on research areas such as IoT application to digital manufacturing, programming for production plans, IoT-based intelligent sensor systems, IoT architectures, IoT thinking and principles, supervisor control and data acquisition, operational analysis by IoT software, smart network applications and IoT simulators, and applications of programming in order to improve the future performance of PI 4.0 -IoT-CN/AL and CE.
Mraz et al. in 2017 advised scholars to work on the development of bench and site acceptance testing techniques, mathematical modeling, development of architectures for improving plant production and control, safety and controlling of operations in oil refineries, web servers and database information systems, internet technological design, monitoring technologies, energy controlling, multiagent systems, discharge and architecture loss control system, control systems for accidents and failures, scheduling programming, design and application of physical or soft controllers, improvement in industrial network, algorithm/programming configuration of plant system, modeling of hybrid internet and intranet, monitoring the gasification processes, plant control principles and advanced technology design, and application of physical controllers for improving the future performance of PI 4.0 -PC-CN/AL and CE.
Ahmed and Kim [67] advised scholars to concentrate on such techniques as smart network, application for Table 9: Tabulation of URL/DOI of research manuscripts associated with exclusion parameters.    Integrated biofuel filtration and presented biofiltration process for tertiary treatment of oil refinery wastewater. It is suggested that integration of advanced oxidation processes is best for oil refinery and for reducing the wastewater.   Implemented neural networks technique to analyze the behavior of an oil production system and to determine the optimal values of gas injection rate and oil rate lifted from a production system. Implemented the neural networks, linear and sequential programming to analyze the behavior of an oil production system to find the optimal values of gas injection rate and oil rate in a two-oil-well system.    Proposed a design of electronic controller for controlling the air flow arrangement in ventilation and air-conditioning systems in the paint industry.

Neural network technique application 15 Wireless Communications and Mobile Computing
Design and application of physical or soft controllers 19 Wireless Communications and Mobile Computing Proposed few plant control principles to improve the capacity of the produced water treatment (PWT) in offshore oil and gas production processes. Developed an evolutionary algorithm for controlling the quality of ink with respect to three concerns, i.e., microgeometry parametric analysis, analyzing the coverage, and the strength. Suggested three kinds of technologies and strategies, i.e., wait and see, in-process focused, and all round strategies to be used by further industries for diminishing GHG emissions.
Technologies and strategies 20 Wireless Communications and Mobile Computing  Insights of evolutionary algorithms approach are illustrated towards tackling the oil and gas industry. Eventually, it resulted in the shifting of the interest of geosciences community to algorithm applications towards maintaining oil and gas fields.      network, multiagent model to network nodes, multiagent system network, weighted oil trade network, intelligent sensor network application, network for incident reduction, and fuzzytechnique-based network in order to enrich the future performance of PI 4.0 -N-CN/AL and CE.

Application of evolutionary algorithms 21 Wireless Communications and Mobile Computing
Triantafillou [78] emphasized to scholars the importance of focusing on areas such as technologies and strategies, resilience-based modeling, application of strategy perfor-mance measurement, programming, machine learning integration, standard principles and procedures and application of techniques/methods and fuzzy classification modeling, accident practice-based model development, application of evolutionary algorithms, analysis-based vibration monitoring, international electrotechnical commission protocols, benefits of logistic system for oil industry, literature survey report, general management, technical architectures, feasibility analysis of processes, antitheft system application, production management, stochastic frontier analysis, functional network intelligent clustering system, data envelope Agrifoglio et al. [36] stimulated the scholars to focus on areas such as the effects of digital technologies on optimization of multiple operations, high-level architecture simulation, theoretical knowledge, operation management by hydrogen networks, illustration of industrial tools and techniques, infrared spectroscopy applications, illustration of cogeneration applications, regenerated spent catalyst scheme applications, and group decision-making algorithms to improve the future performance of PI 4.0 -IDM-CN/AL and CE.

Managerial Implication/Research Values for Scholars
The presented research work suggested the moderate and very weak performing research areas corresponding to PI 4.0 -CN/ALs. This work advises researchers to accept as a research gap the very weak performing research areas of PI 4.0 and to focus on them to enhance the future performance of PI 4.0 -CNs/ALs, linked to CE. The research work also provides a new research methodology to PI 4.0 researchers for materializing the future RTs of PI 4.0 under multiple/different CNs/ALs. Researchers can avail the same methodology to address the future RTs and improve CE. The work has iconic value if it can be conducted without using any bibliographic software tools.

Conclusions
Investigation has shown that strong CNs/ALs lead a vital role in driving the processes of PIs at a faster and quicker rate. Machine learning, big data analysis, signal analysis of sensors, machine to machine virtual interaction, and mechanical automation thrust PI 4.0 to turnout the standard/predicted outputs for attaining CE. Strong CNs/ALs aid PI 4.0 in controlling the quality of refined beverage items, escalating green practices, and stimulating the overall sustainable traceability of PSs. In the presented research forum, the authors built a PI 4.0 -CN/AL model by gratifying RQ 1 and archiving 302 research documents on conducting SLS over 2007-2017; however 275 were respected under the inclusion parameters to represent the RTs of the CN/AL model by satisfying RQ 2 . Later, the RTs of the presented CN/AL model materialized by addressing RQ 3. It has been concluded that the RTs of a particular DO is dazzling among defined PI 4.0 -CNs/ALs. The DO-RT has been found with consistent acceleration and momentum. RTs of residue CNs/ALs are expressed in descending orders, i.e., N>D>A>PC>IoT/VR>I>CPS/ER-P/IDM (discussed in Section 5). The authors also bifurcated RTs under two aspects, where N>D>A>PC>IoT/VR>I are introduced under moderately weak performing research areas/CNs/ALs, whereas CPS/ERP/IDM are introduced under very weak performing research areas/CNs/ALs. In continuation of above, the authors suggested which areas the scholars should focus on to reform and amend the RT's level of moderate and very weak performing CNs/ALs (discussed in Section 5.1) thus hiking and improving CE. The research work can aid future scholars with methodology to materialize the RTs of any interdisciplinary research area and topic focusing on CE. The presented CN/AL model also assists PI 4.0 researchers and managers to explore the same model for investigating and mapping the performance of PI 4.0 by using expert's opinion/subjective data with focus on CE aspects.

Data Availability
Highlights: this work proposes a novel method to measure and identify the growths and trends of IIoT communication networks for petrochemical companies