Received January 19, 2021, accepted February 1, 2021, date of publication February 22, 2021, date of current version March 2, 2021.Digital Object Identifier 10.1109/ACCESS.2021.3060863The Role of AI, Machine Learning, and Big Data inDigital Twinning: A Systematic Literature Review,Challenges, and OpportunitiesM. MAZHAR RATHORE 1, (Member, IEEE), SYED ATTIQUE SHAH2, DHIRENDRA SHUKLA3,ELMAHDI BENTAFAT1, AND SPIRIDON BAKIRAS 1, … Continue reading “A Systematic Literature Review | My Assignment Tutor”
Received January 19, 2021, accepted February 1, 2021, date of publication February 22, 2021, date of current version March 2, 2021.Digital Object Identifier 10.1109/ACCESS.2021.3060863The Role of AI, Machine Learning, and Big Data inDigital Twinning: A Systematic Literature Review,Challenges, and OpportunitiesM. MAZHAR RATHORE 1, (Member, IEEE), SYED ATTIQUE SHAH2, DHIRENDRA SHUKLA3,ELMAHDI BENTAFAT1, AND SPIRIDON BAKIRAS 1, (Member, IEEE)1Division of Information and Computing Technology, College of Science and Engineering, Hamad Bin Khalifa University, Doha, Qatar2Data Systems Group, Institute of Computer Science, University of Tartu, 51009 Tartu, Estonia3Dr. J. Herbert Smith Centre, University of New Brunswick, Fredericton, NB E3B 5A3, CanadaCorresponding author: Spiridon Bakiras (sbakiras@hbku.edu.qa)ABSTRACT Digital twinning is one of the top ten technology trends in the last couple of years, due to its highapplicability in the industrial sector. The integration of big data analytics and artificial intelligence/machinelearning (AI-ML) techniques with digital twinning, further enriches its significance and research potentialwith new opportunities and unique challenges. To date, a number of scientific models have been designedand implemented related to this evolving topic. However, there is no systematic review of digital twinning,particularly focusing on the role of AI-ML and big data, to guide the academia and industry towards futuredevelopments. Therefore, this article emphasizes the role of big data and AI-ML in the creation of digitaltwins (DTs) or DT-based systems for various industrial applications, by highlighting the current state-ofthe-art deployments. We performed a systematic review on top of multidisciplinary electronic bibliographicdatabases, in addition to existing patents in the field. Also, we identified development-tools that can facilitatevarious levels of the digital twinning. Further, we designed a big data driven and AI-enriched referencearchitecture that leads developers to a complete DT-enabled system. Finally, we highlighted the researchpotential of AI-ML for digital twinning by unveiling challenges and current opportunities.INDEX TERMS Digital twin, artificial intelligence, machine learning, big data, industry 4.0.I. INTRODUCTIONDigital twinning is a process that involves the creation of avirtual model (i.e., a twin) of any physical object, in orderto streamline, optimize, and maintain the underlying physical process. Theoretically, the digital twin concept was firstpresented in 2002 by Grieves et al. [1] during a specialmeeting on product life-cycle management at the Universityof Michigan Lurie Engineering Center. In his subsequentarticle [2], he further defined digital twinning as a combination of three primary components: 1) a virtual twin; 2) acorresponding physical twin (a physical object that can bea product, a system, a model, or any other component suchas, a robot, a car, a power turbine, a human, a hospital, etc.);and 3) a data flow cycle that feeds data from a physical twinto its virtual twin and takes back the information and processes from the virtual twin to the physical twin. The virtualThe associate editor coordinating the review of this manuscript andapproving it for publication was Claudio Zunino .twin is nothing but an algorithm that replicates the behavior(fully or partially) of the corresponding physical counterpart,by generating the same output as does the physical object ongiven input values. Mostly, it is considered as part of the smartmanufacturing process, but it can be used in any domain, suchas construction, education, business, transport, power andelectronics, human and healthcare, sports, and networkingand communications.Digital twinning was first adopted by Tuegel et al. [3]in 2011 to digitally reproduce the structural behavior of anaircraft. Initially, digital twinning was used as a maintenance tool to continuously monitor the craft’s structure. Then,it was replicated as a complete twin in order to simulateits entire life-cycle and predict its performance [3]. Later,digital twinning started gaining popularity in several industries that aimed at making their processes smarter, intelligent,and optimally dynamic, based on the operating conditions.The technology raises its global demand, as it facilitates infinding the product flaws, reducing production cost, real-time32030 This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ VOLUME 9, 2021M. M. Rathore et al.: Role of AI, Machine Learning, and Big Data in Digital Twinning: A SLR, Challenges, and Opportunitiesmonitoring of resources, and increasing the life of the productby predicting product failure. On this account, digital twinning became one of the top-ten technology trends [4].Several surveys have been published, highlighting the current research trends of digital twinning in various fields.For instance, Wanasinghe et al. [5] pointed out the state-ofthe-art works of digital twinning in the oil and gas industries. Lu et al. [6] and Cimino et al. [7] reviewed thecurrent reference models, applications, and research issuesin manufacturing. Qi and Tao [8] emphasized on the role ofdata and digital twinning in achieving smart manufacturing.DT-related patents are discussed by Tao et al. [9] in differentindustries. And, the modeling perspective of digital twinningis explored by Rasheed et al. [10].Recently, the use of IoT, big data, and AI-ML technologieshave brought new potentials in digital twinning. The adoptionof these techniques ensures a perfect digital twin and introduces new research challenges and opportunities. Since 2015,several digital twins have been developed in various industries using AI-ML and big data analytics, and the number ofrelated research articles is growing rapidly. Despite the growing popularity, adaptability, and applicability of AI-enableddigital twinning in the industrial sector, exploited by IoTand big data technologies, no systematic review has beenperformed that explicitly focuses on the role of these technologies in digital twinning. The above-mentioned surveysdo not fully cover the importance of these technologies in theDT domain. Therefore, there is an exigency of a systematicapproach towards the thorough review of the current developments in AI-enabled digital twinning using IoT technologyand big data. This can drive both academia and industrytowards further research, by highlighting the current findings,future potentials, challenges, and applications of AI-enableddigital twinning in the industrial sector.In this article, we carried out a systematic literature reviewthat incorporates all the research work in the form of articles,patents, and web-reports, covering digital twinning and itsintegration with state-of-the-art AI-ML and big data analyticstechniques. We highlighted the role of big data, AI, machinelearning, and IoT technologies in the process of digital twincreation, by listing examples from current deployments invarious industrial domains. We introduced the digital twinparadigm, by explaining its basic concepts and highlightingits applications in several industrial areas. After a thoroughliterature survey, we identified 1) tools that can be used fordigital twin creation; 2) the criteria for successful digital twinning; and 3) research opportunities and challenges in digitaltwinning for diverse industrial sectors. Finally, we designeda reference model for digital twinning that exploits IoT, bigdata, and AI-ML approaches.The rest of the paper is organized as follows. Section IIbriefly presents the survey methodology. Section III formallydefines digital twinning, its creation method, and other basicconcepts. Section IV summarizes the application of digitaltwinning in various industries. Section V briefly describesbig data and AI, while Section VI discusses the relationshipbetween IoT, big data, AI, and digital twinning. Section VIIsummarizes the role of AI in digital twinning with stateof-the-art research developments. Section VIII outlines theimportant data-driven patents in digital twinning. Section IXpresents the evaluation criteria for an ideal digital twinning, and Section X lists the tools that may be required inthe process of digital twinning. The design details of thereference architecture for AI-enabled DT creation is presented in Section XI, while the current research opportunitiesand research challenges in digital twinning are described inSection XII. The article is concluded in Section XIII.II. METHODOLOGYTo the best of our knowledge, the survey at hand is thefirst of its kind in terms of reviewing AI-ML and big dataanalytics techniques for digital twinning. The systematic literature review (SLR) carried out in this study is based onthe guidelines recommended by [11], [12], with the aim ofsummarizing the current literature and establishing the basisfor qualitative synthesis and information extraction. SLR isan organized, efficient, and widely recognized method thatis comparatively better than the traditional literature reviewprocess [13].We identified the following six research questions thatdirected our entire review processV1) What is digital twinning, how does it work, and whatare the standards and technologies to create a digitaltwin (DT)?2) What is the relationship between AI-ML, big data, IoT,and digital twinning?3) What is the role of AI-ML and big data analytics indigital twinning, its related applications, and currentdeployments in different industrial sectors?4) What are the tools required for the creation ofAI-enabled DT?5) What is the criteria for a successful DT or DT-basedsystem?6) What are the main challenges, market opportunities,and future directions in digital twinning?To capture the wide range of digital twinning applications, we searched eight multidisciplinary electronic bibliographic databases, including 1) IEEE Xplore (IEEE, IET);2) ACM digital library; 3) Scopus (ScienceDirect, Elsevier); 4) SpringerLink (Springer); 5) Hindawi; 6) IGI-Global;7) Taylor & Francis online; and 8) Wiley online library.We also searched the US patents database. Using suitablesearch strings is crucial to extracting the appropriate literature from the electronic bibliographic databases. Due tothe diverse nature of this study, we used a set of appropriate keywords that assures the inclusion of AI-ML andbig data analytics in industrial digital twinning. Specifically,as shown in Table 1, we defined various keywords, combinedwith logical operators, to search the electronic bibliographicdatabases.The search was carried out just before August 2020. Priorto 2015, we found very few papers on digital twinning.VOLUME 9, 2021 32031M. M. Rathore et al.: Role of AI, Machine Learning, and Big Data in Digital Twinning: A SLR, Challenges, and OpportunitiesTABLE 1. Search strings.FIGURE 1. Number of journal papers published by different libraries.In 2017, the topic gained popularity and became one of thetop 10 trends in strategic technology [14]–[16]. In the period2015–2020, more than 2000 Scopus-indexed journal articles,more than 1000 patents, 250 book chapters, and 20 bookshave been published, discussing digital twinning technology.However, we identified over 850 articles that match the searchcriteria defined in Table 1. Fig. 1 and 2 show the totalnumber of journal and conference papers published on thetopic of digital twinning by the different libraries. Amongother publishers, IGI-Global published seven articles, Hindawi published three articles, and ACM published only twoarticles in their journals. Additionally, Fig. 3 illustrates the piechart of published articles related to various applications ofDT (it includes both conference and journal papers). Clearly,manufacturing is the dominant application area for digitaltwinning.Considering the aforementioned research questions,we defined a set of inclusion and exclusion criteria for anarticle as followsV1) The study is written in English.2) The study is published in a scientific journal, magazine,book, book chapter, conference, or workshop.3) The journal article is included only if the journal’simpact factor is > 1:0.4) The conference article is included only if the conference is mature enough (it has already published at leastfifteen versions of its proceedings).5) Publications such as dissertations, in-progress research,guest editorials, poster sessions, and blogs areexcluded.6) Duplicate papers that appear in several electronicdatabases will only be considered once.7) The study is excluded if not fully focusing on the digitaltwinning concept or any of its specified applications.FIGURE 2. Number of conference papers published by different libraries.FIGURE 3. AI-ML driven digital twinning research statistics in differentfields.Among the 850 articles that matched the designated keywords, a total of 213 papers were selected after applyingthe above inclusion and exclusion criteria. IEEE ACCESSand Elsevier Journal of Manufacturing Systems are the toptwo journals that have published the most articles within theset criteria. The selected publications were first evaluated onthe basis of their titles and abstracts. The concept of digitaltwinning in relation to the research questions was criticallyexamined, and a total of 63 papers were excluded in thisphase. Some paper-abstracts were not clear enough to bedirectly evaluated, hence a full-text screening was performedon 150 papers, resulting in the exclusion of 52 additionalpapers. Snowball sampling was performed on the remainingset of 98 papers. Then, we used the references and citations ofthe selected papers to perform backward and forward search,respectively, for identifying new potential papers.Finally, a total of 117 papers concerning digital twinning, its applications, and related technologies, were selectedfor data extraction and synthesis of this study. Among the117 articles, 61 articles discussed AI-ML based digital twins.For each selected article, metadata forms were maintainedto categorize the information about the articles and to notethe observations assessed. The extracted metadata was thencoded for analysis, according to the year of publication,authors’ names, affiliated universities or organizations, keywords, name of journal or conference, research model, areaof focus, data source, and opportunities/issues highlighted.The categories were derived according to the data needed to32032 VOLUME 9, 2021M. M. Rathore et al.: Role of AI, Machine Learning, and Big Data in Digital Twinning: A SLR, Challenges, and Opportunitiesanswer the research questions and for identifying the paper’smain research areas. In addition to journal and conferencearticles, we included 20 US patents, 15 technical web-reports,and 5 standards, focusing on digital twinning. Some otherarticles that indirectly relate to digital twinning, such as supporting tools, technologies, and survey methodologies, arealso referred in our study.III. DIGITAL TWIN: INTRODUCTION AND BACKGROUNDA. DIGITAL TWIN: DEFINITIONS AND CONCEPTResearchers define digital twins in several ways. The pioneersof digital twinning, Grieves and Vickers [17], define a digitaltwin as ‘‘a set of virtual information constructs that fullydescribes a potential or actual physical manufactured productfrom the micro atomic level to the macro geometrical level.At its optimum, any information that could be obtained frominspecting a physical manufactured product can be obtainedfrom its Digital Twin.’’ In their opinion, the digital twin canbe any of the following three types: 1) digital twin prototype(DTP); 2) digital twin instance (DTI); and 3) digital twinaggregate (DTA). A DTP is a constructed digital model ofan object that has not yet been created in the physical world,e.g., 3D modeling of a component. The primary purpose of aDTP is to build an ideal product, covering all the importantrequirements of the physical world. On the other hand, a DTIis a virtual twin of an already existing object, focusing on onlyone of its aspects. Finally, a DTA is an aggregate of multipleDTIs that may be an exact digital copy of the physical twin.For example, the digital twins of a spacecraft structure and aspacecraft engine are considered DTIs that may be aggregatedinto a DTA.In this article, we assume the concepts of DTI and DTAwhen referring to a DT. Note that, the majority of academicscholars and industries follow similar definitions for a digitaltwin. For instance, Glaessgen and Stargel [18] defined it fromthe perspective of vehicles as ‘‘A digital twin is an integrated multiphysics, multiscale, probabilistic simulation ofan as-built vehicle or system that uses the best available physical models, sensor updates, fleet history, etc., to mirror the lifeof its corresponding flying twin.’’ Similarly, Tao et al. [19]considered the aspect of product life cycle and interpretedthe digital twin as ‘‘a real mapping of all components in theproduct life cycle using physical data, virtual data and interaction data between them.’’ Söderberg et al. [20] focused onthe application of optimization while defining a digital twin.According to them, digital twinning is an approach to performa real-time optimization to a physical system using its digitalcopy. Finally, Bacchiega [21] made it simpler by defining itas ‘‘a real-time digital replica of a physical device.’’With our understanding, shown in Fig. 4, digital twinningis a process that involves the construction of 1) a cybertwin that digitally projects a living or non-living physicalentity or a process (a system); and 2) a physical connectionbetween cyber and physical twins to share data (and information) between them aimed at dynamic optimization, real-timemonitoring, fault diagnostics and early prediction, or healthFIGURE 4. Digital twinning concept.monitoring of the physical counterpart. A physical twin canbe a process, a human, a place, a device, or any other objectwith a special purpose, and which is able to be replicated inthe digital world as either a partial twin with limited functionalities, or a complete twin that incorporates the full behaviorof its physical peer. Digital twinning is mostly employedin industries for physical objects in their units. However,there exist some digital twins that are mirrors of processesin the physical world, such as digital twins of a mobile-edgecomputing (MEC) system [22], human protein–protein interaction (PPI) [23], supply chain [24], components-assembly ata manufacturing unit, and job scheduling [25].B. A BRIEF HISTORY OF DIGITAL TWINSThe idea of creating a digital copy of a physical entity wasintroduced in the early 2000s. However, the term ‘‘digitaltwin’’ originated around ten years ago. Michael Grieves,in one of his articles [2], claimed that the concept of digital twins was first presented during a lecture on productlife-cycle management (PLM) in 2003. Whereas, in his otherbook chapter [1], he stated that the concept was originallyproposed, without a name, in 2002 while presenting a paperin a special meeting at the University of Michigan LurieEngineering Center. Grieves mentioned in this book chapter,‘‘While the name has changed over time, the concept andmodel has remained the same.’’ He added that it was giventhe name ‘‘mirrored spaces model (MSM)’’ in 2005 andchanged to ‘‘information mirroring model’’ in 2006. NASAstarted using this concept of virtual and physical models intheir technology roadmaps [26] and proposals for sustainable space exploration [27] since 2010. However, the name‘‘digital twin’’ was first coined in 2011 by John Vickersof NASA. Practically, the first digital twin was developedby Tuegel et al. [3] for the next-generation fighter aircraft,in order to predict its structural life.C. OPERATIONAL MECHANISMAlthough the digital twin concept was introduced in 2002,it became a popular trend due to the advancement in sensortechnology and IoT, which play a vital role in digital twinningby collecting real-time data from the physical world andsharing it with the digital world. The twinning can be viewedas a bridge between a physical twin and the correspondingVOLUME 9, 2021 32033M. M. Rathore et al.: Role of AI, Machine Learning, and Big Data in Digital Twinning: A SLR, Challenges, and Opportunitiesvirtual twin. The physical-to-virtual connection is establishedwith a technology that allows the transfer of information fromthe physical environment to its virtual twin, including webservices, cellular technology, WiFi, etc. The virtual twin isadjusted gradually with the functioning of the physical twinby continuously collecting the differences between the twoenvironments. These connections allow the monitoring ofresponses to both conditions and interventions. The conditions mainly occur in the physical environment, whereas theinterventions take place within the virtual twin. Thus, a digitaltwin holds a real-time status of the physical counterpart.The virtual-to-physical connections represent theinformation circulating from the virtual to the physical environment. This information may change the state of the physical twin by displaying some data or changing the system’sparameters (for optimization, diagnostics, or prognostics).Although virtual-to-physical connections are very helpful inDT modeling, they are not always included in the description.Instead, it is common to consider a one-way connection,i.e., physical-to-virtual. Finally, the data and the informationfrom both physical and virtual worlds are stored and analyzedat a centralized server—or a cloud computing platform—where the final decisions related to optimization, diagnostics,or prognostics, are made.D. DIGITAL TWIN STANDARDSCurrently, there is no particular standard that solely focuseson the technical aspects of digital twinning. Standardization efforts are under-development by the joint advisorygroup (JAG) of ISO and IEC on emerging technologies [28].However, the ISO standard ISO/DIS 23247-1 [29] is theonly standard that offers limited information on digital twins.In addition, there are other related standards that may facilitate DT creation. For example, the ISO 10303 STEP standard [30], the ISO 13399 standard [31], and the OPC unifiedarchitecture (OPC UA) [32] technically describe ways toshare data between systems in a manufacturing environment.IV. DIGITAL TWINNING IN INDUSTRIES: APPLICATIONSDigital twinning is becoming apparent in various industries,including manufacturing, medical, transportation, business,education, and many more. In this section, we present the roleof digital twinning and the current research followed in theseareas.A. MANUFACTURINGDigital twinning is conceived as a major tool in the manufacturing industry to carry out smart manufacturing, faultdiagnosis, robotic assembly, quality monitoring, job shopscheduling, and meticulousness management. In this way,Rosen et al. [33] emphasizes the use of digital twinning inmanufacturing. Modules in a computerized system communicate with each other during every step of the production, thusdepicting a realistic model of its physical counterpart. Similarly, the work by Qi and Tao [8] explains the benefits of bigdata-driven DT in smart manufacturing. The DT combinesall the manufacturing processes, starting from product designto maintenance and repair. The virtual model is capable ofidentifying the constraints of the virtual design in the physicalworld, which are iteratively improved by the designers. Dataproduced by sensors and IoT devices are then analyzed andprocessed using big data analytics and AI applications toenable the manufacturers to select a satisfactory plan.On the other hand, DT is also used to monitor a componentor a product, considering its usage, health, and performanceduring the life-cycle of manufacturing. Real-time data provided to the virtual model allows it to self-update and predictany abnormal behaviors. Optimal solutions are developed forproblems found in the virtual models, and the actual manufacturing model is adjusted accordingly. Maintenance and repairof the physical system can also be scheduled timely, based onthe predictions of the virtual models. One of such digital twinprojects is originated by Slovak University of Technology inBratislava [34] for a physical production line of pneumaticcylinders, where they defined the continuous optimization ofproduction processes and performed proactive maintenance,based on the real-time monitoring data. Similarly, a digitaltwin of manufacturing execution system (MES) was developed by Negri et al. [35] that enables the supervisory control over the physical MES system using sensor technology,by allowing the multi-directional communication betweendigital and physical sides of manufacturing assets.Several state-of-the-art works highlight that DTs shouldbe capable of self-healing and predictions. These predictionsplay a vital role in an important aspect of smart manufacturing, i.e., fault diagnosis, since a minor issue during productioncan cause irreparable damages. A variety of technologiesused in fault diagnosis like Support Vector Machines [36],Bayesian Networks [37], Deep Learning [38]–[40], and manyothers [41]–[44] are capable of enhanced fault diagnosis.However, Xu et al. [45] highlight that, in production systems, conditions are constantly changing. Therefore, the sametraining model cannot be applied throughout the process, butcreating a new model requires a lot of time and resources.As such, they proposed a digital twin-assisted fault diagnosis using deep transfer learning (DFDD) approach. DFDDhas been applied to fault diagnosis in smart and complexmanufacturing. The framework involves two phases. In thefirst phase, the virtual model of the system is constructed.Repeated designs of the model are tested and evaluated in thevirtual space until all anomalies are discovered. Simulationdata during design testing is provided to an embedded faultdiagnosis model in the virtual space. The diagnosis modelkeeps learning from the simulation data using Deep Neural Networks, in order to increase its efficiency for failureprediction during the start of the production phase whenthere is insufficient training data. The second phase startswhen the virtual model achieves acceptable performance. Thephysical entity is constructed and linked to its correspondingvirtual model. Data is transferred from a physical entity to thevirtual model through sensors during production. A diagnosismodel is formed and updated using the current data from32034 VOLUME 9, 2021M. M. Rathore et al.: Role of AI, Machine Learning, and Big Data in Digital Twinning: A SLR, Challenges, and Opportunitiesthe physical entity and the knowledge learned from theprevious phase, which is transferred using deep transferlearning (DTL).Robotic assembly, in industrial manufacturing, is responsible for handling a notable amount of work [46]. It is involvedin packaging, labeling, painting, welding, and many others.With the advancements in the complexity of manufacturing, these robotic assemblies have become more error-prone.The concept of DT is being utilized in this area to monitorand optimize the assembly process. In [47], a multisourcemodel-driven digital twin system (MSDTS) is designed forrobotic assembly. The MSDTS model consists of three parts.The physical space consists of sensors, its associated data,and the robotic arm for moving and gripping objects. Thevirtual space consists of a server, a multisource model, and avirtual reality display and control (VRDC). A communicationinterface offers the exchange of data between two spaces inreal-time. Initially, a 3D model of the entire physical spaceis constructed using a depth sensor that is mounted on therobot arm. During operation, the VRDC provides a completeview of the physical system by receiving a video stream froman RGB camera. When the robot arm moves, angular data issent to the virtual twin through the communication interfacein real-time, and the graphical model in the virtual systemfollows the same trajectory. The physical contact of the robotarm with the surrounding object is simulated in the virtualsystem using the Kelvin-Voigt model (KVM), where parameters of the model are estimated through the data of contactforce and relative motion of contact point. A surface-baseddeformation algorithm is used to simulate the deformationof an object using the data generated by KVM. The resultsof the models are rendered in the VRDC. A complete viewof the system is provided to the operator via a head mount.Interaction with the physical space is done using a controlhandle.Another important element in manufacturing is job shopscheduling, which makes efficient use of resources toreduce production time and maximize production efficiency. In real-life situations, due to errors and anomalies,the scheduling process can be rendered inefficient. With theintroduction of smart manufacturing and digital twins, newDT-based job shop scheduling methods are introduced toovercome scheduling plan deviation and provide a timelyresponse. One such model is proposed in [48]. A DT-basedjob shop consists of a physical and a virtual space, whichcommunicate through a CPS. Scheduling data from the physical space is sent to the virtual space, and multiple schedulingstrategies are simulated and retrieved from the virtual models.The finalized scheduling plan is fed into the physical space.Since a physical system has many modules, the plan is dividedand categorized based on the respective modules. Continuouscommunication between the physical and virtual space resultsin achieving precise scheduling parameters, as well as prediction of any disturbances in the schedule. The schedulingplan can hence be updated and fed to the physical system forincreased efficiency and timely response.Digital twin and big data are playing an important rolein smart manufacturing starting from product life-cycle tomaintenance and repair. Some of the stated research articleshighlighted the importance of digital twinning in the areasof smart manufacturing. The concept of utilizing a variety ofdata and integrating it with IoT, virtual reality, and data analytics, results in high fidelity monitoring, timely predictionand diagnosis of faults in assembly or production, and overalloptimization and improvement of the manufacturing process.B. MEDICALApplications of DT in medical include the maintenance ofmedical devices and their performance optimization. DT,along with AI applications, are also used to optimize thelife-cycle of hospitals by transforming a large amount ofpatient data into useful information. The ultimate aim of thedigital twinning in healthcare is to help authorities in managing and coordinating patients. Mater private hospitals inDublin (for cardiology and radiology) were facing problemsregarding increased services, patient demand, deterioratingequipment, deficiency of beds, increased waiting time, andqueues. These problems indicated the call for the improvement in the current infrastructure to cater to increasingneeds. Mater private hospitals (MPH) partnered with SiemensHealthineers to develop an AI-based virtual model of theirradiology department and its operations [49]. As a result,the simulations of the model provided insights towards theoptimization of workflows and layouts. The realistic 3D models of the radiology department, provided by DT techniques,allowed for the prediction of operational scenarios and theevaluation of the best possible alternatives to transform caredelivery.In recent years, with the introduction of ‘‘precisionmedicine,’’ the focus of DT technology is shifted towards ahuman DT. Precision medicine is the branch of healthcarethat promotes tailored treatments on an individual level. Thehuman DT would be linked to its physical twin and woulddisplay the processes inside the human body. It can resultin an easier and accurate prediction of illness with propercontext, and bring a paradigm shift in the way patients aretreated. Virtual physiological human (VPH) was the earliesthuman DT that was developed [50]. VPHs would act as a‘‘Virtual Human Laboratory’’ where each VPH was modifiedbased on the specific patient, and different treatments wouldbe tested on the modified VPH platform.Apart from human DTs, organs or human body parts digitaltwins have also been developed. Data from Fitbit devices,smartphones, and IoT devices are sent in real-time to suchDTs, in order to provide constant feedback regarding humanorgan activity. Some organs’ DTs have been used by expertsto perform clinical analysis, whereas many others are underdevelopment. In a study, a 3D digital twin of a heart wasdeveloped by Siemens Healthineers [51], after performinga comprehensive research on approximately 250 millionimages, functional reports, and data. The model exhibitedthe physical and electrical structure of a human heart. ThisVOLUME 9, 2021 32035M. M. Rathore et al.: Role of AI, Machine Learning, and Big Data in Digital Twinning: A SLR, Challenges, and OpportunitiesDT is currently under research at the Heidelberg universityhospital (HUH), Germany, where DTs of 100 patients havebeen created, who had a history of heart diseases within aperiod of six years. Simulations of these DTs were comparedwith the ground truth, which provided promising results.Another DT of the heart has been developed by researchersat the Multimedia Communications Research Lab in Ottawa,Canada. It is called a Cardio Twin and targets the detectionof ischemic heart disease (IHD) [52]. IHD is a conditioncharacterized by reduced blood flow to the heart, which canlead to chest pain or mortality in case of delayed treatment.The researchers developed the DT on the concept of edgecomputing/analytics, where the time is considered very critical. Data is collected from social networks, sensors, and medical records. The accumulated data is fed to an AI-inferenceengine, where data fusion, formatting, and analytics are performed using TensorFlow Lite to discover new information.The Cardio Twin can communicate with the physical twin inthe real world, using a multimodal interaction component thatemploys WiFi/4G or Bluetooth communication. Cardio Twinperformed a sample classification of 13420 ECG segmentswith an accuracy of 85.77%, in a short span of 4.84 seconds.However, no method to evaluate Cardio Twin in the real worldhas been introduced.Sim&Cure, a company based in Montpellier, France,developed a simulation model for the treatment of aneurysm.Aneurysm is an outward bulging of blood vessels, typicallycaused by an abnormally weakened vessel wall. A seriouscase of aneurysm can result in clotting, strokes, or death.The last option for treating aneurysm is surgery. However,endovascular repair (EVAR) is generally used, since it isless invasive and low-risk. In EVAR, a stent-graft/catheteris placed into the affected area to minimize the pressure.In many cases, choosing the stent-graft/catheter is difficultand depends on the size of the blood vessels. The Sim&Cure’sDT helps surgeons in selecting an ideal implant to cater to thesize of the aneurysm as well as the blood vessels. A 3D modelof the affected area and surrounding vessels is created, andmultiple simulations are run on the personalized DT, whichallows surgeons to have a better picture. Promising resultshave been presented in preliminary trials [53], [54].Researchers at the Oklahoma State University developed ahuman airway DT—named ‘‘virtual human’’—in their computational biofluidics and biomechanics laboratory (CBBL).They tracked the flow of air particles in aerosol-deliveredchemotherapy and found out that, the aerosol-based drughit the cancerous cells with less than 25% accuracy [55].This caused more harm than benefits to patients, as theremaining drug would fall on healthy tissue. The version1.0 of ‘‘virtual human’’ was based on a 47-year-old standingmale, containing the entire respiratory system. V1.0 alsoallowed patient-specific structural modifications, e.g., creating a respiratory system of a standing/seated female or a kidwith/without respiratory conditions. Following the success ofV1.0, CBBL researchers developed its successor version 2.0.The V2.0 was patient-specific, and was created by performingan MRI/CT scan of the patient. The scanned data was usedto construct a 3D model of the lungs. The researchers atCBBL then created a virtual population group (VPG), whichwas a large group of human DTs. The VPG exhibited trendswithin different groups/sub-groups. Simulations to analyzethe trends of aerosol particle movement were conducted onthe VPG, by varying the particle sizes, inhalation rate, andinitial position of the medication. These simulations indicatedthat the drug’s effectiveness would increase to 90% if the drugdelivery method was personalized to each patient, rather thandistributing the drug evenly for every patient [55].In another study, Liu et al. [56] proposed a cloud-basedDT healthcare solution (CloudDTH) for elderly people. Thecloud-based solution provides a fusion of physical and virtualsystems to address real-time interaction between patientsand medical institutions, and personalized healthcare for theentire life-cycle of the elderly. CloudDTH has a layered architecture, providing health resources, identification of medicalpersonnel, user interface, virtualization, and security servicesto users. CloudDTH obtains real-time data from sensors forECG, BP, pulse rate, and body temperature. These sensorsare already implemented in the CloudDTH framework. Thesensor data are then transmitted to the cloud server, usingTCP. In case of an incident, such as patient falling, heartattack, stroke, etc., the monitoring model, after performinganalysis on the received data, sends a high-frequency andmulti-attribute monitoring order of the patient to medicalpersonnel. A case study was performed by researchers, wheredata from two patients with normal and abnormal heart rateswas input to the system. The simulation results indicatedsymptoms of arrhythmia in one patient, and recommendedthe dosage of medication based on their physical conditions.The CloudDTH framework simulations also provided a feasible scheduling mechanism for elderly patients in hospitals,in order to avoid long queues.C. TRANSPORTATIONNumerous innovative technologies have been brought forward with the development of IoT, including digital twins,autonomous things, immersive technology, etc. Various typesof digital twins are developed in transportation sector, including DTs for automobile components, vehicles, vehicularnetworks, and road infrastructures. However, the purposeremains the same i.e., monitoring, optimization, and prognostics and health management. For example, Wang et al. [57]developed a framework for connected vehicles based ondigital twins. The framework used vehicle-to-cloud (V2C)communication to provide advisory speed assistance (ADSA)to the driver. Real-time data from sensors was obtained inthe physical system, which was sent to the cloud throughthe V2C module. All processing of the data from V2C wasperformed on the cloud server. The computed results weresent back to the physical system and served as a guidance system for components within the physical world. The authorsdemonstrated the effectiveness of their framework with a casestudy of cooperative ramp merging involving three passenger32036 VOLUME 9, 2021M. M. Rathore et al.: Role of AI, Machine Learning, and Big Data in Digital Twinning: A SLR, Challenges, and Opportunitiesvehicles, and the results showed that the digital twin canindeed assist transportation systems.Cioroaica et al. [58] worked on the context of connectedvehicles in smart ecosystems. The establishment and achievement of goals in smart ecosystems are possible when smartentities within the ecosystem co-operate with each other.This is achieved when the systems have a level of trust.The authors developed a virtual hardware-in-the-loop (vHiL)testbed model to evaluate the trust-building capability ofsmart systems within an ecosystem. A smart agent, capable of interacting with the vehicle’s electronic control unit(ECU), is installed at the vehicle along with its correspondingDT. In Phase 1, the trustworthiness of the smart agent isevaluated by simulation in its corresponding virtual twin.Phase 2 involves trust-building, where the smart agent isexecuted on the ECU. Evaluation of simulated and actualresults identifies the obstacles. These obstacles are overcomeby the collaboration of virtual and physical entities to achievetrustworthiness in a smart ecosystem.Chen et al. [59] studied the use of unmanned aerial vehicles (UAVs) as complementary computation resources in amobile edge computational (MEC) environment for mobileusers (MU). MEC provides computational capabilities toMUs within a radio access network (RAN). Mobile users sendcomputational tasks to UAVs by creating the correspondingVMs. The tasks arriving at the UAVs are stored in queues and,due to limited resources, the MUs have to compete for them.The authors proposed deep reinforcement learning (DRL)techniques for the scheduling of tasks on the UAV, and forminimizing the response delay from the UAV to the MUs.The training of the DRL network in an offline manner isachieved by creating a digital twin of the entire MEC system.Simulations with varying parameters were conducted and thebest results were selected. The results of the DRL schemetrained on digital twins ensured significant performance gainswhen compared to other baseline approaches.Digital twins have also been utilized in transportation systems for traffic congestion management, congestion prediction, and avoidance. Kumar et al. [60] worked on intelligenttransport systems, leveraging technologies such as fog/edgeanalytics, digital twins, machine learning, data lakes, andblockchain. The authors captured situational informationfrom cameras, and performed edge analytics on the acquireddata. An entire virtual vehicle model was created via a digital twin to replicate the real-world scenario. Driver intentions were predicted using machine and deep learning algorithms to avoid traffic congestion. This virtual vehicle modelallowed autonomous vehicles to make decisions regardingoptimal paths, but also helped drivers of non-autonomousvehicles to make better decisions based on the traffic situationand the mined driver intentions.Digital twins have also been used for the maintenanceof different systems. The work implemented by Venkatesan et al. [61] monitored and projected the health conditionsof electric motor vehicles using an intelligent digital twin(i-DT). The framework tracked the health of the electricmotor in an electric vehicle using fuzzy logic and artificialneural networks (ANNs). The average speed of the vehicleand the duration of travel was fed into the ANN i-DT andfuzzy logic i-DT for training purposes. In addition, simulations carried out on a digital twin tested the performance ofthe entire framework. Parameters such as winding and casingtemperature, deterioration in magnetic flux, and lubricantrefill time were set for the digital twin. The comparison oftheoretical and i-DT computations indicated that an i-DT caneffectively be used in electric vehicles to foresee their health.D. EDUCATIONAnother important area where digital twins can play a crucialrole, is education. Digital twins of physical entities suchas labs, construction, mechanical equipment, can be createdand provided to students for online learning. However, therehas not been a lot of research effort on the use of DT inthe education domain. One such work was performed bySepasgozar [62] that used digital twins and virtual gamingfor online education. The authors created a digital twin ofan excavator along with a virtual game for the course ofconstruction management and engineering. The project contained four modules named 1) group wiki project and roleplay (GWiP); 2) interactive construction tour 360 (ICRT 360);3) virtual tunnel boring machine (VTBM); and 4) pilingaugmented reality and digital twin (PAR-DT). GWiP wasused for doing group projects online. ICRT 360 consisted ofrecorded videos to provide details on construction sites andmachinery. VTBM was a virtual game-based environmentthat helped students to learn about the working of a tunnelboring machine. Virtual equipment was introduced in thegame, where a student or a group of students could exploretheir interests. PAR was developed for smartphones and Oculus headsets to provide students an augmented environmentto collaborate and understand the importance of piling inconstruction. The final module involved a digital twin ofan excavator, which was also linked to a physical instance.The DT provided hands-on learning about the functions andmovements of an excavator. The students’ feedback emphasized the importance of an immersive environment in onlineeducation.E. BUSINESSBusiness is also one of the areas where DT is playing animportant role. According to PropTechNL [63], the real estatesector is fragmented in terms of architects, installation, construction, transport, and management. This fragmentationresults in an inefficient system that has a negative impacton people living in a society. Digital twins can provide hugeopportunities in real state, and facilitate the creation of smartsocieties. For example, a wide range of sensors can collectdata, and the performance of a building can be measured andimproved. Digital twins in real estate may add significantvalue by re-positioning buildings to the needs and requirements of customers, hence improving the customer experience. The design of buildings, the usage, effectiveness, andVOLUME 9, 2021 32037M. M. Rathore et al.: Role of AI, Machine Learning, and Big Data in Digital Twinning: A SLR, Challenges, and Opportunitiesstrength of raw materials, as well as maintenance and runningcosts, can be managed through digital twins. Thus, it providesa cost-effective, fast, and smart way of developing a realestate. For instance, an American multinational company,GE Healthcare, has incorporated the use of DT to redesignits systems, in order to run new hospitals more efficiently.Kampker et al. [64] introduced a framework for the development of successful business models in smart services. Thescenario of crop (potato) harvesting was taken into consideration during their research. In traditional harvesting mechanisms, the harvesting machines are set up based on historicaldata and the experiences of individual operators. However,the lack of standard procedures may cause damage to thecrop. Therefore, the authors developed a framework, basedon a digital twin, to reduce the damage to the crop duringharvesting. Specifically, a digital twin is set up near the physical field. The virtual model then passes through the samestages as the real crop. During the simulation, the conditionof the neighboring crop is analyzed for potential damage.The results of the analysis lead to adjusting the parameters,and repeated simulations continue until the optimal settingsare found. Tests carried out by the authors indicated thatmore damage to the crop is caused by its impact on multipleconveyor belts during the transition. Hence, adjustment tothe height and position of conveyor belts can reduce therisk of damage. This framework can also tweak the settingsof autonomous harvesting machines, apart from providingrecommendations to operators.F. OTHER INDUSTRIESDigital twinning can be a part of smart construction, wherea DT may be designed for buildings, roads, or any otherinfrastructure development. For example, a virtual twin wasdeveloped for office buildings [65] that manages the building’s life-cycle, by collecting data through sensors. Furthermore, DT technology may advance the disaster managementapproaches in smart cities [66]. Possibly, the technology alsohas a potential to protect industrial control systems and datafrom cyber attacks. On this account, Dietz and Pernul [67]proposed the use of digital twinning technology to identifysecurity threats that target industrial control systems (ICSs),and rectify their effects. Theoretically, they focused on theStuxnet worm [68] that compromised the speed of centrifuge,and Triton [69] that digitally invaded a petrochemical plantin Saudi Arabia. Deitz et al., indicated in the Stuxnet example that the outliers in the historical network traffic wouldhave detected a threat. Similarly, in the case of simulations,the deviation of network traffic between the virtual and physical systems would have identified the attack.V. AI-ML AND BIG DATA: AN INTRODUCTIONBig data remains one of the top research trends from lastfew years. It is different from an ordinary data because ofits high volume, high velocity, and heterogeneous variety,as interpreted in Fig. 5. Researchers named these characteristics as ‘‘the 3Vs of big data,’’ i.e., volume, velocity, andFIGURE 5. Big data definition.variety. Later, two more Vs—value and veracity—were addedto the list. Thus, we refer to any data as big data, if it isof significant size (volume), it is being produced at veryhigh-speed (velocity), and it is heterogeneous with structured,semi-structured, or unstructured nature (variety). The worthof big data analytics brings the fourth V (i.e., value) intoits characteristics, thus making it an asset to the organization. Big data analytics is a process that analyzes big dataand converts it to valuable information, using state-of-the-artmathematical, statistical, probabilistic, or artificial intelligence models. However, the 3Vs of big data lead us to a newworld of challenges, including capturing, storing, sharing,managing, processing, analyzing, and visualizing such highvolume, high-velocity, and diverse variety of data. To this end,various frameworks [70]–[73] have been designed to handlebig data for effective analytics in different applications.Artificial intelligence (AI) is the digital replication ofthree human cognitive skills: learning, reasoning, and selfcorrection. Digital learning is a collection of rules, implemented as a computer algorithm, which converts the realhistorical data into actionable information. Digital reasoningfocuses on choosing the right rules to reach a desired goal.Whereas, digital self-correction is the iterative process ofadopting the outcomes of learning and reasoning. Every AImodel follows this process to build a smart system, whichperforms a task that normally requires human intelligence.Most of the AI systems are driven by machine learning, deeplearning, data mining, or rule-based algorithms, where othersfollow logic-based and knowledge-based methods. Nowadays, machine learning and deep learning are widely used AIapproaches.It is often confusing to differentiate between artificialintelligence, machine learning, and deep learning techniques.Machine learning is an AI method, which searches for particular patterns in historical data to facilitate decision-making.The more data we collect, the more accurate is the learningprocess (reflects the value of big data). Machine learningcan be 1) supervised learning, which accepts data sets with32038 VOLUME 9, 2021M. M. Rathore et al.: Role of AI, Machine Learning, and Big Data in Digital Twinning: A SLR, Challenges, and Opportunitieslabeled outputs in order to train a model for classification orfuture predictions; 2) unsupervised learning, which works onunlabeled data sets and is used for clustering or grouping; and3) reinforcement learning, which accepts data records with nolabels but, after performing certain actions, it provides feedback to the AI system. Examples of supervised learning techniques are regression, decision trees, support vector machines(SVMs), naive Bayes classifiers, and random forests. Similarly, K-means and hierarchical clustering, as well as mixture models, are examples of unsupervised learning. Finally,Monte Carlo learning and Q-learning fall under the reinforcement learning category. On the other hand, deep learning is amachine learning technique that is motivated by biologicalneural networks with one or more hidden layers of digitalneurons. During the learning process, the historical data areprocessed iteratively by different layers, making connections,and constantly weighing the neuron inputs for optimal results.In this article, we mainly focus on digital twin systems basedon machine learning.VI. RELATIONSHIP BETWEEN IoT, BIG DATA, AI-ML,AND DIGITAL TWINSThe emerging sensor technologies and IoT deployments inindustrial environments have paved the way for several interesting applications, such as real-time monitoring of physical devices [74], indoor asset tracking [75], and outdoorasset tracking [76]. IoT devices facilitate the real-time datacollection—that is necessary for the creation of a digital twinof the physical component [77], [78]—and enable the optimization [79] and maintenance [80] of the physical component by linking the physical environment to its virtual image(using sensors and actuators). Note that, the above-mentionedIoT data is big in nature [81] (as explained in Section V),so the big data analytics can play a key role in the development of a successful digital twin. The reason is that industrial processes are very complex, and identifying potentialissues in early stages is cumbersome, if we use traditionaltechniques. On the other hand, such issues can easily beextracted from the collected data, which brings efficiencyand intelligence into the industrial processes. However, handling this enormous amount of data in the industrial and DTdomains requires advanced techniques, architectures, frameworks, tools, and algorithms. For instance, Zhang et al. [82],[83] proposed a big data processing framework for smartmanufacturing and maintenance in a DT environment.Oftentimes, cloud computing is the best platform for processing and analyzing big data [84]. Additionally, an intelligent DT system can only be developed by applying advancedAI techniques on the collected data. To this end, intelligenceis achieved by allowing the DT to detect (e.g., best process strategy, best resource allocation, safety detection, faultdetection) [85], predict (e.g., health status and early maintenance) [80], [86], optimize (e.g., planning, process control, scheduler, assembly line) [87], [88], and take decisionsdynamically based on physical sensor data and/or virtual twindata. In short, IoT is used to harvest big data from the physicalFIGURE 6. Relationship between IoT, big data, AI-ML, and digital twins.environment. Later, the data is fed to an AI model for thecreation of a digital twin. Then, the developed DT can beemployed to optimize other processes in the industry. Theoverall relationship among IoT, big data, AI, and digital twinsis presented in Fig. 6.VII. CURRENT DEPLOYMENTS OF DIGITAL TWINS USINGBIG DATA AND MACHINE LEARNINGWe have identified the primary sectors where DT-based systems are developed with the help of AI-ML techniques. In thefollowing sections, we discuss the current deployments inthese sectors, including smart manufacturing, prognosticsand health management (PHM), power and energy, automotive and transport, healthcare, communication and networks,smart cities, and others.A. SMART MANUFACTURINGSmart manufacturing involves 1) the acquisition of data frommanufacturing cells through a variety of sensors; 2) themanagement of the acquired data; and 3) the data exchangebetween different devices and servers. In a DT environment,the data is collected from a physical manufacturing cell and/orits corresponding virtual cell. Such data can be further utilizedfor manufacturing process optimization, efficient assemblyline, fault diagnosis, etc., using AI approaches. The AI-MLbased digital twinning process for smart manufacturing isdepicted in Fig. 7.Manufacturing is the top industry where most digital twinsare being developed. Xia et al. [91] proposed a manufacturingcell digital twin to optimize the dynamic scheduler for smartmanufacturing. An intelligent scheduler agent, called digitalengine, was developed and trained for optimization usingdeep reinforcement learning algorithms (DRLs), such as natural deep Q-learning [101], double deep Q-learning [102],and prioritized experience replay (PER) [103]. The underlying features were captured from both the physical and virtualenvironments of the cell by an open platform communications (OPC) server. The training of the DRL network wasdone through a gradient descent process, which requires finitelearning iterations and is sufficiently intelligent, reliable, androbust. The developed DT-based dynamic scheduler optimizes the manufacturing process by accelerating the training,testing, and validation of smart control systems. The systemwas tested on a robot cell to optimally select the strategyVOLUME 9, 2021 32039M. M. Rathore et al.: Role of AI, Machine Learning, and Big Data in Digital Twinning: A SLR, Challenges, and OpportunitiesFIGURE 7. DT-based smart manufacturing using big data analytics andAI-ML.for performing the lower level tasks that are necessary toaccomplish the higher level manufacturing goal.Zhou et al. [79] performed a geometric optimizationof centrifugal impeller (CI) by collecting features, suchas meridional section (MS), straight generatrix vectors(SGV), and set of streamlines (SSL), from both the physical and CAD-based digital model of the CI. However, withthe improvement in machinability, the DT-based geometricoptimization reduces the aerodynamic performance. Thus,the best model for the CI is selected by training the deepdeterministic policy gradient (DDPG) reinforcement learning model [104] to iteratively select the fair geometry ofthe CI-design with optimum values of machinability andaerodynamic performance. For the DDPG algorithm, theyused two actor networks (online and target network) as thestrategy function π to control the agent-actions, and two criticnetworks (online and target network) to evaluate these actionsand give rewards. The proposed DT-based optimization spedup significantly the design and manufacturing of the impeller.Similarly, Zhang et al. [95] also developed an impeller DT,but for the purpose of manufacturing process planning. Theyemployed a knowledge reuse deep learning network (PKRNet) [105], which takes data from dynamic knowledge base,views from 3D computer-aided impeller design (CAD), 2Ddrawings, and process knowledge. The objective is to optimize the theoretical processes and generate the best processplan for product manufacturing, by considering both manufacturing time and monetary cost.Furthermore, Lee et al. [106] designed a deep learning andcyber-physical system based digital twinning (DTDL-CPS)architecture for smart manufacturing, that can be used in shopfloor optimization, fault diagnosis, product design optimization, and predictive maintenance. BDHDTPREMfg [84] isa similar CPS-based big data-driven model for DT-enabledre-manufacturing. Several other digital twins have beendeveloped in the manufacturing industry using AI approachesthat could not be fully discussed in this article. Rather,Table 2 summarizes these digital twins with respect to theproblem they solved (i.e., the application), the ML-approachthey used to solve the problem, and the DT use-case theydeveloped.B. PROGNOSTICS AND HEALTH MANAGEMENTThe persistent use of a product degrades its performance overtime, which may lead to malfunctioning. Thus, prognosticsand health management (PHM) is very crucial in all industries. PHM process involves the prediction of the remaininguseful life of a product and the consistent monitoring of itshealth. This is the second most important application of DT,following smart manufacturing. Note that, several alternativeterms, such as ‘‘predictive modeling’’ [86], ‘‘structural lifeprediction’’ [3], ‘‘remaining useful life’’, and ‘‘predict andact’’ [107] have also been used in place of PHM. DT-basedPHM regularly monitors the physical equipment based onthe data generated by the equipment-sensors, performs diagnosis and prognosis operations on the data using big dataanalytics and AI, and recommends design rules for immediatemaintenance. The process of DT-enabled PHM is depictedin Fig. 7.Tao et al. [108] developed a digital twin for a wind turbinein a power plant, in order to monitor its health by performing gearbox prognosis and fault detection. The proposedDT-driven PHM can be applied to any complex equipment inharsh environments, such as aircraft, ships, and wind turbines.The wind turbine DT is built based on various geometrylevels, physics, behavior, and rules. The DT can detect thedisturbances in the turbine environment, as well as potentialfaults in itself and its designed model. The data is collectedfrom the DT model (both physical and digital) and is matchedagainst the thresholds for degradation detection. In addition,past DT-data is used to train a single hidden layer neuralnetwork for better prediction of gradual faults and detectionof its causes, using extreme learning machine (ELM) [109].The abrupt fault in the turbine is detected by comparing thedata from the physical and virtual environments. Similarly,to improve ship efficiency and avoid unnecessary maintenance operations, a data-driven ship digital twin was developed by Coraddu et al. [110]. Their goal was to determine thespeed loss due to marine fouling. Multilayered-deep extremelearning (DELM) [111] predicts the ship’s speed, based on thefeatures collected from on-board sensors, such as designedand ground speed, draft, engine and shaft generator power,wind speed, temperature, fuel consumption, etc. The expectedship speed is compared with the measured speed to computethe speed loss. Finally, robust linear regression is applied tothe speed loss information to determine whether the speedloss is due to marine fouling.Numerous other digital twins have been developedfor PHM of industrial components, including photovoltaic energy conversion unit [112], battery system [113], vehicle motor [61], UAV [115], spacecraft [116],32040 VOLUME 9, 2021M. M. Rathore et al.: Role of AI, Machine Learning, and Big Data in Digital Twinning: A SLR, Challenges, and OpportunitiesTABLE 2. State-of-the-art AI-ML developments in digital twinning for smart manufacturing.aircraft [3], [118]–[120], gillnet [122], gearbox, aircraftturbofan engine, rotating shaft-bearing [121], etc. All thesesystems are summarized in Table 3.C. POWER AND ENERGYIn the power and energy sector, most of the DTs are developedin electronic systems, wind-power farms, cooling systems,and fuel-related systems. The digital twin of an invertermodel [125] was developed by imitating the voltage controller, the current control loop, and the controlled plant,based on three distinct neural networks (NNs). Each ofthe three NNs is trained on real data collected from thephysical model, where the back propagation (BP) algorithmis deployed to tune, in real-time, the proportional–integral(PI) controller. Also, Andryushkevich et al. [126] introducedthe digital twin of power-system using ontological modeling. The developed DT selects the optimal configurationof the hybrid power supply system, by utilizing geneticalgorithms [127]. Likewise, a digital twin framework forpower grids was designed by Zhou et al. [128] to perform real-time analysis. Specifically, NN-based learning wasapplied to predict the grid operational behavior for fast security assessment, based on the voltage stability and oscillationdamping.In addition, a DT for a dew-point cooler was developed [99] to improve its cooling performance, by optimizingoperational and design parameters, including cooling capacity, coefficient of performance (COP), dew point efficiency,wet-bulb efficiency, supply air temperature, and surface area.The DT of the cooler is developed with feed-forward neuralnetworks (FFNNs), and digitally mimics the cooler’s behavior by utilizing the air characteristics (i.e., temperature, relative humidity) as well as the main operational and designparameters (i.e., air velocity, air fraction, HMX height, channel gap) as inputs to the FFNN. Later, the DT-collected dataare supplied to a genetic algorithm (GA) for multi-objectiveevolutionary optimization, in order to maximize cooling,COP, and wet-bulb efficiency, and minimize the surface areawithin four diverse climates (i.e., tropical rainforest, arid,Mediterranean hot summer, and hot summer continental climates).Apart from design and performance optimizations,ML-based PHM is accomplished for power and energyrelated components with the use of DTs, such as windturbine, [108], electric vehicle motor [61], photovoltaic systems [112], battery systems [113], plasma radiation detectionin metal absorber–metal resistor bolometer [114], as discussed in Section VII-B.VOLUME 9, 2021 32041M. M. Rathore et al.: Role of AI, Machine Learning, and Big Data in Digital Twinning: A SLR, Challenges, and OpportunitiesTABLE 3. State-of-the-art AI-ML research in industrial digital twinning for PHM.D. VEHICLES AND TRANSPORTATIONA vehicle digital twin was developed by Alam and ElSaddik [85] in a vehicular cyber-physical system (VCPS),by mimicking its speed behavior, fuel consumption, andairbag status. The system utilized fuzzy rule base with aBayesian network [129], in order to build a reconfigurationmodel for driving assistance. Similarly, Kumar et al. [60]built virtual models of running vehicles in the cloud, whichobtained real-time road and vehicular data through fog oredge devices, in order to avoid traffic congestion. The driverbehavior and intention are predicted using machine learningon historical data. LSTM-based recurrent neural networks(RNNs) [130] are applied on the data to obtain the bestroute for a particular vehicle. Besides, digital twins havealso been developed for vehicular network system, itself.For instance, the digital twin of a mobile edge computing (MEC) system was developed [59] for resource allocation in unmanned aerial vehicle (UAV) networks, using deeprecurrent Q-networks (DRQNs) [131]. Likewise, the digitaltwin of software-defined vehicular networks (SDVNs) [132]allows for the predictive verification and maintenance diagnosis of running vehicles network, using machine learning.Furthermore, prognostics and health management is conducted by developing digital twin of aircraft [118] and spacecraft [116], ship [110], and electric vehicle motor [61]. All ofthese PHM approaches employ machine learning techniques.E. HEALTHCAREIn healthcare, the majority of AI-ML enabled DTs are humandigital twins [23], [56], [133]–[136]. Mimicking the fullfunctionalities of a human body is not currently possible,thus, a human digital twin can only focus on limited aspectsof human biology. For example, the digital twin by Barricelli et al. [133] focuses on fitness-related measurements ofathletes. Specifically, their virtual patient classified physical athletes and predicted their behavior using KNN classifiers [137] and support vector networks [138], by trainingmodels on physical patient data collected by IoT devices.Björnsson et al. [23] concentrated on protein–protein interaction (PPI) networks to diagnose and treat patients of aparticular disease. Their model is implemented as an AIsystem that monitors the effect of drugs on the human body,using machine learning tools, such as Bayesian networks,deep learning, and decision trees.Furthermore, Chakshu et al. [135] mimicked the patient’shead behavior for detecting the severity of carotid stenosis.Their model selects components from a patient video andapplies principal component analysis (PCA) to identify theseverity of carotid stenosis, by comparing it with the virtualmodel components. The authors also recommended the use ofdeep learning, machine learning, and other AI techniques forbetter detection accuracy. Similarly, Mazumder et al. [134]digitally replicated the process of generating synthetic PPG32042 VOLUME 9, 2021M. M. Rathore et al.: Role of AI, Machine Learning, and Big Data in Digital Twinning: A SLR, Challenges, and Opportunitiessignals to create the digital twin of a cardiovascular system. In the virtual model, parameters are optimized using aparticle-swarm-optimization (PSO) algorithm. The algorithmminimizes the integral-squared-error (ISE) in the feature set,in order to generate the synthetic PPG signal. On the otherhand, Laamarti et al. [136] and Liu et al.’s [56] models aregeneric ML-enabled frameworks for providing health services to elderly people.F. COMMUNICATIONS AND NETWORKSIn the networking and communications domain, the digital twin of an indoor space environment [139] is implemented to model, predict, and control the terahertz (THz)signal propagation characteristics in an indoor space. TheDT selects the best THz signal path from the base stationto the mobile target, by avoiding obstacles. The DT identifies the obstacle, its position, and dimensions, by applying a you only look once (YOLO) machine learning algorithm [140] on the monochromatic image of the obstacle.Furthermore, deep learning algorithms are used for materialtexture recognition and classification. On the other hand,a new network architecture, equipped with ML-based virtualtwin of a software-defined vehicular network (SDVN) [132],is designed to benefit from intelligent networking and adaptive routing. Dong et al. [22] developed a similar digital twinof a real network for mobile edge computing. The virtualmodel of the MEC is equipped with a deep neural networkthat is frequently updated based on the variation of the realnetwork. The model then selects the optimal resource allocation and offloading policy at each access point.G. SMART CITIESIn the smart city sector, a Zurich city digital twin [141]was developed by transforming 3D spatial data and citymodels, including buildings, bridges, vegetation, etc., to avirtual world. The authors discussed the effects of urban climate, which can be predicted by machine learning techniquesbased on the current weather and air-quality data. Similarly,a Vienna city digital geoTwin [142] can be linked with citydata, such as socioeconomic, energy consumption, and maintenance management data, in order to make it a living digitaltwin with the aim of AI technologies. Furthermore, a visionfor integrating artificial and human intelligence for a disastercity digital twin is introduced by Fan et al. [66]. Finally,a geospatial digital twin [143] is the digital replica of a spatialentity, where machine learning and deep learning techniquesare used for interpretation, analysis, and organization of 3Dpoint clouds.H. OTHER INDUSTRIESDT systems that utilize AI-ML techniques have beendeployed in other industries as well. For instance, the supplychain DT by Marmolejo-Saucedo [24] was developed for apharmaceutical company, using machine learning and patternrecognition algorithms. The objective was to identify thebehavior, dynamics, and changing trends in the supply chain.Data management for DT environments is another area ofactive research. Specifically, a DT-enabled collaborative datamanagement framework was proposed, using edge and cloudcomputing [100]. The goal was to perform advanced dataanalytics in additive manufacturing (AM) systems, in order toreduce the development time and cost, and improve the product quality and production efficiency. To this end, the authorsintroduced cloud-DTs and edge-DTs, developed at differentproduct life-cycle stages, which communicate with each otherin order to support intelligent process monitoring, control,and optimization. As a use case, the framework was implemented within the MANUELA project, where layer defectanalysis was performed by a deep learning model on productlife-cycle data. Moreover, Tong et al. [144] introduced anintelligent machine tool (IMT) digital twin model for machining data acquisition and processing, using data fusion and MLapproaches.VIII. DATA-DRIVEN DIGITAL TWINNING PATENTSThe importance of DT technologies can be verified by thenumber of patents in this field. In particular, more than onethousand patents have been awarded on AI-enabled digital twinning in all around the world. A wind-power farmdigital twin was filed as a U.S. Patent in 2016 by General Electric (GE) [145], where the DT is composed of twocommunication networks: 1) a farm-based communicationnetwork, which enables the coupling of control systems fromindividual wind turbines with the main wind farm controlsystem and with other wind turbines; and 2) a cloud-basedcommunication network that is composed of an infrastructure of digital wind-turbine models, where the plurality ofthe models are continuously changing during farm operation, by investigating and analyzing data generated by thefarm-based communication network using machine learning.Furthermore, they provided a fully functional graphical userinterface (GUI) of the digital wind-turbines, where the usercan control the input features of the DT model to optimizethe performance of the wind farm using machine learningalgorithms. In another patent, Shah et al. [146] developed thedigital twin of a vehicle cooling system, by using status data(such as health scores) to predict cooling system failures andoptimize its performance. Similar data-driven digital twinning systems have been designed in the energy and powersector [147].In predictive analytics for machine maintenance, GE’s Hershey et al. [148] developed a system to predict the lifetime ofa component in the electromechanical industry (such as anaircraft engine), by developing a digital twin of the physicalsystem. The component is monitored by IoT-based sensorsand its remaining life is assessed based on the monitoringconditions. In this process, they developed a stress analysismodel, a fluid dynamics model, a structural dynamic model,a thermodynamic model, and a fatigue cracking model. Then,they utilized probabilistic models, such as a Kalman filter,to predict the lifetime and detect component faults. Similarly, the Siemens corporation designed a generic digitalVOLUME 9, 2021 32043M. M. Rathore et al.: Role of AI, Machine Learning, and Big Data in Digital Twinning: A SLR, Challenges, and Opportunitiestwin model [149] for a variety of machines, including heating, ventilation, and air conditioning (HVAC). They utilized data-driven approaches for energy-efficient machinemaintenance, utilizing sensor data and model-based analytics. Several other patents focus on predictive analytics withAI-enabled digital twins [150]–[152].A few digital twin patents have also been developed in thehealthcare sector. GE researchers designed a patient DT [153]to diagnose diseases, treat, and prescribe medicines. Thedigital representation of the patient (i.e., the DT) consists ofmedical record data structures, medical images, and historicalpatient information. The DT is equipped with healthcare software applications (such as expert systems), patient medicaldata, and AI models (neural networks, machine learning) thatcan diagnose, identify health issues, and prescribe treatments(e.g., medication, surgery, etc.). Also, Nagesh [154] buildan X-ray tube DT to predict tube-liquid bearing failures.He used X-ray tube housing vibration data, collected by asensor in a free run mode of an X-ray tube, and appliedAI-based prediction. There are also patents in DT-basedsurgery for the healthcare industry that utilize data-drivenapproaches [155], [156].Finally, there are hundreds of additional patents thatemphasize AI-enabled data-driven digital twinning, whichcould not be covered here. These digital twinning systemsbelong to a variety of industrial sectors, including manufacturing [157], [158], run-time environment [159], transportand automotive industry [160]–[162], building and construction systems [163], etc.IX. EVALUATING A SUCCESSFUL DIGITAL TWINA successful digital twin can only be justified when its virtualtwin closely matches the functionality of its physical counterpart. This justification can be presented by comparing theoutputs of the physical and virtual models, and computingthe loss. On this account, accuracy is the main factor toconsider when evaluating digital twins. On the other hand,the purpose of building a digital twin also matters in evaluating its success. This can be justified by the performanceimprovement of the corresponding physical system that isattributed to its digital twin. For example, for a DT whosepurpose is to optimize the assembly line, the improvementcan be measured by computing the number of actions (or subtasks) and the time taken to manufacture a full component(or to complete a main task/goal) with the DT and withoutDT. This is also the case with other applications, including product design optimization, product performance optimization, process optimization, control optimization, scheduler optimization, resource management, component PHM,etc. In addition, the processing time and efficiency of thedigital twinning system can also be one of the successcriteria.In addition, when using AI or machine learningapproaches, the accuracy of the selected model affects theefficacy of the DT. Specifically, the accuracy of the underlying ML model, as well as the feature selection process andthe amount of training data, may greatly affect the outcomeof the DT. Therefore, when designing a DT-based systemthat employs ML techniques, we have to select the modelwith the higher accuracy and efficiency. The same approachshould be taken with the selection of other technologies forDT-development, such as IoT, edge computing, and cloudcomputing.To this end, only a few state-of-the-art digital twinning systems have been fully evaluated in the literature. For instance,Zhang et al. [87] assessed their job-floor digital twin bycomparing the performance of the job-floor with and withoutdigital twinning. They selected job scheduling time, utilityrate, and job tardiness as performance parameters. Similarly,Zhang et al. [93] highlighted the importance of digital twinning by showing the performance improvement in processtime, fault time, and maintenance time of blisk machining dueto its digital twin. Likewise, Min et al. [164] conveyed a risein the oil yield ratio due to a petrochemical industry DT. Furthermore, Xu et al. [45] used the accuracy of fault diagnosisas a metric to assess the performance of the developed virtualtwin. Finally, Akhlaghi et al. [99] verified the accuracy of thedeveloped twin by comparing the outputs of the digital andphysical twins. They also showed the effectiveness of theirdigital twinning mechanism, by pointing out the optimizationachieved for the dew point cooler. All the aforementionedDTs were developed using various machine learning modelsand, in each case, the authors selected the model that providedthe best accuracy.X. DIGITAL TWIN DEVELOPMENT TOOLSThere is no standalone technology for DT implementation, rather, there is an integration of multiple technologies,including big data, AI-ML, IoT, CPS, edge computing, cloudcomputing, communication technologies, etc. Every technological component can be implemented with a varietyof tools. Here, we only focus on the tools that facilitatecomponents integration, digital twin simulation, twins bridging, physical twin control, data storage and processing, andmachine learning. Table 4 presents the summary of widelyused tools that may provide support in different stages ofdigital twinning.Integrating physical components for data collection andthen digitally mimicking them in a virtual environment aretwo important stages of digital twinning. There are varioustools available to accomplish these tasks in an industrial unit.Siemens MindSphere is one of the widely used tools to integrate components in a manufacturing industry. Siemens alsodeveloped an object-oriented-based Tecnomatix API to simulate physical components in a virtual environment, as usedby [91]. The Open Simulation Platform (OSP) is another one,which is jointly developed by the Det Norske Veritas Germanischer Lloyd group (DNV GL), the Norwegian University of Science and Technology (NTNU), Rolls-Royce, andSINTEF Ocean. OSP can digitally mimic any component ofthe maritime industry. Other popular integration and simulation tools are FIWARE, Predix (a cloud-based platform from32044 VOLUME 9, 2021M. M. Rathore et al.: Role of AI, Machine Learning, and Big Data in Digital Twinning: A SLR, Challenges, and OpportunitiesTABLE 4. Digital twinning supporting tools.VOLUME 9, 2021 32045M. M. Rathore et al.: Role of AI, Machine Learning, and Big Data in Digital Twinning: A SLR, Challenges, and OpportunitiesGE digital), CNC machine tools control platform IndraMotion MTX, Beacon, Thingworx, and others.Next, bridging physical and virtual twins is another primary aspect of digital twinning. This bridge is used by a virtual twin to harvest the real-time data from the correspondingphysical peer using sensors. On the other side, the physicalpeer is controlled (optimized) based on the output of thevirtual twin. Popular tools in the market to facilitate the bridging between physical and virtual twins are TwinCat, SAP,Codesys, CNC tools, Aspera, and RaySync. Similarly, thereare few applications that are used in initial modeling and twindesign, such as ANSYS Twin Builder, MWorks, Siemens NXsoftware, SolidWorks, Autodesk tools, and FreeCAD.In the machine learning domain, there are hundreds ofmodels available for tasks such as optimization, prediction,classification, and clustering. However, there is no singleplatform that offers APIs for all existing ML models. Themost widely used and well-known libraries for implementing, training, and testing supervised ML-models are Tensorflow, CNTK and Caffe. Keras and Weka provide easier anduser-friendly interfaces for developing basic machine learning models. There are also commercial tools available, suchas Mathworks Matlab, which is equipped with vast librariesof neural networks and Microsoft-Azure implemented MLmodels. Reinforcement learning is one of the most populartechniques that is widely used for dynamic optimization andprocess planning in digital twinning. To this end, OpenAI’sGym and rllab are tools with standardized interfaces forreinforcement learning.Industrial components produce large amounts of data,termed as big data, which are hard to process with standarddata management tools in a digital twin environment. Hadoopis one of the most popular ecosystems for big data processingthat offers parallel processing capabilities with multiple compute nodes. Apache has also developed several tools for bigdata processing and effective analysis, including Cassandra,Spark, Storm, S4, Hive, Mahout, Flink, and HBase. Mostof the Apache tools are open-source and support machinelearning APIs. Similar tools include HPCC by LexisNexisRisk Solution, Qubole, Statwing, Pentaho, and VoltDB.XI. DATA-DRIVEN REFERENCE ARCHITECTURE FORDIGITAL TWINNINGTo effectively exploit the value-added capabilities offered bythe integration of big data analytics and AI-ML within thescope of digital twinning, we present a novel reference modelderived from the conducted systematic literature review.Fig. 8 shows the designed reference layered-architecture forthe efficient handling of big data analytics in DT-based industrial environments. The process starts with the collection ofdata from the physical environment (using sensors and actuators) or from the virtual environment (using computer-aidedsoftware and/or simulations). The data is fed to the data analysis and decision-making layer, where AI models, statisticaland probabilistic approaches, or mathematical models areemployed to create the DT-based system or the digital twinitself. During the entire process, various big data processingtools may be utilized, such as Hadoop, Storm, S4, Spark,etc., that allow for parallel processing on multiple computenodes. Fig. 9 depicts the overall data flow for creating anML-enabled digital twin, and then using it for optimization,PHM, or other purposes. First, the virtual model is createdby deploying one of the AI models on the data generated bythe physical twin. Once the digital twin is produced, the datafrom both the physical and virtual twins are given to otherAI models to achieve the given industrial goals, such asdesign optimization, dynamic process planning, healthcare,or PHM. Moreover, the results can be further used to updateand improve both the physical and virtual twins.XII. MARKET OPPORTUNITIES AND RESEARCHCHALLENGESA. MARKET OPPORTUNITIES AND RESEARCH AREASBased on the detailed literature survey, we have summarizedthe following major application areas where DT research canplay a vital role.1) OPTIMIZATIONOptimization is required in almost every industrialprocess, including product design, product performance, process planning, assembly line, task-scheduling, and resourceallocation. Digital twinning is an emerging technology thatprovides a direct pathway to optimization with little effort.However, careful consideration of the optimization algorithm(i.e., ML model) and the underlying feature set (for theoptimization algorithm) is desired for better results.2) PROCESS MONITORING, DIAGNOSTICS, ANDPREDICTIONDigital twins can be developed for industrial process monitoring, defect diagnosis (i.e., product quality assurance),dynamic process or product design updating for time andcost savings, industrial process surveillance (e.g., robot DTfor obstacle avoidance), product time-to-complete prediction,and damage detection.3) PREDICTIVE ANALYTICS FOR MANUFACTUREDPRODUCTSThe quality of every physical entity degrades over time, thusaffecting its performance. Early detection of failures maypromote on-time maintenance, fatigue avoidance, as well astime and cost savings. Such failures can be attributed to faultsand cracks in the product, performance degradation due toaging, and other minor or major complications. Moreover,health monitoring is crucial for certain components that maypotentially cause human casualties, e.g., brake systems incars, vehicles, aircraft, and ship engines, fueling systems,gearboxes, etc. Digital twinning is the most powerful technology for predictive analytics and health monitoring of physicalcomponents. This is also an area where AI-ML techniquescan have a significant impact.32046 VOLUME 9, 2021M. M. Rathore et al.: Role of AI, Machine Learning, and Big Data in Digital Twinning: A SLR, Challenges, and OpportunitiesFIGURE 8. Data-driven reference architecture for digital twinning.FIGURE 9. Overall data-flow framework for digital twinning using bigdata analytics and AI-ML.4) HEALTHCAREDigital twinning has a wider scope in the healthcare sectorwhere human-DTs assist in day-to-day human fitness andhealth monitoring, early disease diagnosis, and the overall well-being of individuals, especially for the elderly andinfants. In addition, it can be used for the treatment orsurgery of patients, by developing a patient-DT. Developingdigital twins for human organs or biological systems willbring a revolution in the healthcare sector, such as DTs forlungs, liver, pregnant female womb or uterus, cardiac system,digestion system, neural system, reproductive system, etc.Other than biological digital twins, the healthcare sector canbenefit by developing DTs for hospitals, medical and surgicalinstruments, remote surgery, surgical processes, etc.5) SMART CITIESIn the context of smart cities, DT technologies can be implemented for traffic systems, smart homes and devices, parking, buildings, livestock, lighting systems, and renewableenergy. Furthermore, 3D virtual city models may facilitateurban planning and monitoring in various smart city areas,including road monitoring and construction, city garbagemanagement, bridge and housing constructions, etc.6) OTHER APPLICATIONSResearch opportunities are not limited to the above-mentionedsectors, but the potential is there in every field, including education, construction, mining, communications and networks,food and agriculture, sports, and so on.B. RESEARCH CHALLENGES AND ISSUESThe rapidly increasing DT popularity and scope, as well as theinvolvement of IoT, big data, and AI technologies, broadenthe research challenges of digital twinning. These challengesare categorized in the following five areas.1) DATA COLLECTIONIoT facilitates data harvesting from a physical twin (usingsensors), data integration, and data sharing with the corresponding virtual twins. This process can amount to a considerable cost. Sometimes, the digital twin may be more costlythan the asset itself, in which case it does not make sense tocreate the DT. On the other hand, the collected data is large(big data), heterogeneous in nature, unstructured, and noisy.Thus, further processing on the data is required to ensureits effective use. Specifically, we need to apply data cleaning techniques, and also organize, restructure, and make thedata homogeneous. Furthermore, controlling the flow of suchlarge amount of data is also a significant challenge. Finally,to improve the accuracy of the DT model, the underlyingVOLUME 9, 2021 32047M. M. Rathore et al.: Role of AI, Machine Learning, and Big Data in Digital Twinning: A SLR, Challenges, and Opportunitiesmachine learning algorithms require a certain amount of datafor training purposes.2) BIG DATA CHALLENGESThe explosive growth of IoT technologies in the industrialsector has led to the generation of large amounts of monitoring (sensor) data. To this end, big data analytics requiresadvanced architectures, frameworks, technologies, tools, andalgorithms to capture, store, share, process, and analyze theunderlying data. There is also a potential for edge and cloudcomputing platforms to handle DT-related data. Specifically,edge computing enables the distributed processing at the network’s edge, while the aggregate processing is accomplishedin the cloud. However, the aggregation of data in the cloudmay cause an increase in response time.3) DATA ANALYSISAI-algorithms for data analytics played a major role in DTfor decision-making, as discussed in the literature. However, the selection of a particular model among hundreds ofML-models with customized configuration is challenging.Every AI-approach has diverse accuracy and efficiency levelswith different applications and datasets (feature set). On theother hand, accuracy may affect the efficiency on the otherside. Hence, depending on the motive and application ofa DT, the selection of the best ML-algorithm and featuresis challenging. Besides, fewer practical implementations ofAI-techniques for digital twinning in the literature raisesmore challenges.4) DT STANDARDIZATION CHALLENGESEven though many digital twins have been developed in various industries, the creation of a complex and reliable digitaltwin demands standardization. Currently, there is no singlestandard that solely focuses on digital twinning. The ISO/DIS23247-1 standard [29] has only limited information on digitaltwinning and, therefore, DT deployment challenges grow dueto the lack of standardization. Standardization efforts areunderway by the joint advisory group (JAG) of ISO and IECon emerging technologies [28].5) SECURITY AND PRIVACY ISSUESSome DT systems, such as human-DTs, product PHM,or defense-related DTs, are considered critical and mayrequire stringent security and privacy guarantees. First, dueto the involvement of IoT devices in digital twinning, a lotof emphasis has to be placed on the security of the underlying communication protocols. Additionally, the large collection of asset-related data needs to be stored securely,in order to prevent data breaches from insider and outsiderthreats.XIII. CONCLUSIONWe performed a systematic literature review of the state-ofthe-art DT systems that employ machine learning and AItechnologies. In particular, we focused on papers publishedin top multidisciplinary electronic bibliographic and patentlibraries, and summarized the current DT deployments ina variety of industries. With the immersion of AI-ML andbig data, digital twinning is evolving at a rapid rate and,with it, a lot of unique challenges and new opportunities areemerging. This article highlighted the research challengesand potentials in many diverse areas, for both academia andindustry. Furthermore, we identified the DT criteria and toolsthat aid its successful development. Finally, we designed areference model for an AI-ML and big data-enabled digitaltwinning system to further guide industrial developers inestablishing DTs that can make their systems smarter, intelligent, and dynamically adaptable to changing conditions.REFERENCES[1] M. W. Grieves, ‘‘Virtually intelligent product systems: Digital andphysical twins,’’ Complex Syst. Eng., Theory Pract., pp. 175–200,2019.[2] M. Grieves, ‘‘Digital twin: Manufacturing excellence through virtualfactory replication,’’ White Paper, 2014, pp. 1–7, vol. 1.[3] E. J. Tuegel, A. R. Ingraffea, T. G. Eason, and S. M. Spottswood,‘‘Reengineering aircraft structural life prediction using a digital twin,’’Int. J. Aerosp. Eng., vol. 2011, pp. 1–14, Aug. 2011.[4] D. Cearley, B. Burke, D. Smith, N. Jones, A. Chandrasekaran, and C. Lu,‘‘Top 10 strategic technology trends for 2020,’’ Gartner, Stamford, CT,USA, Tech. Rep., 2019.[5] T. R. Wanasinghe, L. Wroblewski, B. K. Petersen, R. G. Gosine,L. A. James, O. De Silva, G. K. I. Mann, and P. J. Warrian, ‘‘Digital twinfor the oil and gas industry: Overview, research trends, opportunities, andchallenges,’’ IEEE Access, vol. 8, pp. 104175–104197, 2020.[6] Y. Lu, C. Liu, K. I.-K. Wang, H. Huang, and X. Xu, ‘‘Digital twindriven smart manufacturing: Connotation, reference model, applicationsand research issues,’’ Robot. Comput.-Integr. Manuf., vol. 61, Feb. 2020,Art. no. 101837.[7] C. Cimino, E. Negri, and L. Fumagalli, ‘‘Review of digital twinapplications in manufacturing,’’ Comput. Ind., vol. 113, Dec. 2019,Art. no. 103130.[8] Q. Qi and F. Tao, ‘‘Digital twin and big data towards smart manufacturing and industry 4.0: 360 degree comparison,’’ IEEE Access, vol. 6,pp. 3585–3593, 2018.[9] F. Tao, H. Zhang, A. Liu, and A. Y. C. Nee, ‘‘Digital twin inindustry: State-of-the-art,’’ IEEE Trans. Ind. Informat., vol. 15, no. 4,pp. 2405–2415, Apr. 2019.[10] A. Rasheed, O. San, and T. Kvamsdal, ‘‘Digital twin: Values, challenges and enablers from a modeling perspective,’’ IEEE Access, vol. 8,pp. 21980–22012, 2020.[11] B. Kitchenham and S. Charters, ‘‘Guidelines for performing systematicliterature reviews in software engineering,’’ Keele Univ., Durham Univ.,Keele, U.K., Tech. Rep. EBSE 2007-001, 2007.[12] B. Kitchenham, O. P. Brereton, D. Budgen, M. Turner, J. Bailey, andS. Linkman, ‘‘Systematic literature reviews in software engineering—A systematic literature review,’’ Inf. Softw. Technol., vol. 51, no. 1,pp. 7–15, Jan. 2009.[13] C. Okoli and K. Schabram, ‘‘A guide to conducting a systematic literature review of information systems research,’’ SSRN, Tech. Rep., 2010.[Online]. Available: http://dx.doi.org/10.2139/ssrn.1954824[14] D. Cearley, B. Burke, S. Searle, and M. Walker, ‘‘Top 10 strategic technology trends for 2017: A gartner trend insight report,’’ Gartner, vol. 23,Jun. 2017, Art. no. 6595640781. [Online]. Available: https://www.gartner.com/doc/3645332[15] D. Cearley, B. Burke, S. Searle, and M. J. Walker, ‘‘Top 10 strategictechnology trends for 2018,’’ Gartner, 2017.[16] D. Cearley and B. Burke, ‘‘Top 10 strategic technology trends for 2019,’’Gartner, 2018.[17] M. Grieves and J. Vickers, ‘‘Digital twin: Mitigating unpredictable, undesirable emergent behavior in complex systems,’’ in TransdisciplinaryPerspectives on Complex Systems. Cham, Switzerland: Springer, 2017,pp. 85–113.32048 VOLUME 9, 2021M. M. Rathore et al.: Role of AI, Machine Learning, and Big Data in Digital Twinning: A SLR, Challenges, and Opportunities[18] E. Glaessgen and D. Stargel, ‘‘The digital twin paradigm for future NASAand US air force vehicles,’’ in Proc. 53rd AIAA/ASME/ASCE/AHS/ASCStruct., Struct. Dyn. Mater. Conf., 20th AIAA/ASME/AHS Adapt. Struct.Conf., 14th AIAA, 2012, p. 1818.[19] F. Tao, F. Sui, A. Liu, Q. Qi, M. Zhang, B. Song, Z. Guo, S. C.-Y. Lu, andA. Nee, ‘‘Digital twin-driven product design framework,’’ Int. J. Prod.Res., vol. 57, no. 12, pp. 3935–3953, 2019.[20] R. Söderberg, K. Wärmefjord, J. S. Carlson, and L. Lindkvist, ‘‘Towarda digital twin for real-time geometry assurance in individualized production,’’ CIRP Ann., vol. 66, no. 1, pp. 137–140, 2017.[21] G. Bacchiega, ‘‘Creating an embedded digital twin: Monitor, understandand predict device health failure,’’ Inn4mech-Mechatronics Ind., vol. 4,2018.[22] R. Dong, C. She, W. Hardjawana, Y. Li, and B. Vucetic, ‘‘Deep learning for hybrid 5G services in mobile edge computing systems: Learnfrom a digital twin,’’ IEEE Trans. Wireless Commun., vol. 18, no. 10,pp. 4692–4707, Oct. 2019.[23] B. Björnsson, C. Borrebaeck, N. Elander, T. Gasslander,D. R. Gawel, M. Gustafsson, R. Jörnsten, E. J. Lee, X. Li, S. Lilja,D. Martínez-Enguita, A. Matussek, P. Sandström, S. Schäfer,M. Stenmarker, X. F. Sun, O. Sysoev, H. Zhang, and M. Benson,‘‘Digital twins to personalize medicine,’’ Genome Med., vol. 12, no. 1,pp. 1–4, Dec. 2020.[24] J. A. Marmolejo-Saucedo, ‘‘Design and development of digital twins:A case study in supply chains,’’ Mobile Netw. Appl., vol. 25, no. 6,pp. 2141–2160, Dec. 2020.[25] C. Zhuang, J. Liu, and H. Xiong, ‘‘Digital twin-based smart production management and control framework for the complex productassembly shop-floor,’’ Int. J. Adv. Manuf. Technol., vol. 96, nos. 1–4,pp. 1149–1163, Apr. 2018.[26] R. Piascik, J. Vickers, D. Lowry, S. Scotti, J. Stewart, and A. Calomino,‘‘Technology area 12: Materials, structures, mechanical systems, andmanufacturing road map,’’ NASA Office Chief Technol., 2010.[27] P. Caruso, D. Dumbacher, and M. Grieves, ‘‘Product lifecycle management and the quest for sustainable space exploration,’’ in Proc. AIAASPACE Conf. Expo., Aug. 2010, p. 8628.[28] JETI. Which Technologies is Jeti Considering? Accessed: May 8, 2020.[Online]. Available: https://jtc1info.org/technology/advisory-groups/jeti/[29] Automation Systems and Integration Digital Twin Framework forManufacturing—Part 1: Overview and General Principles, Standard ISO/DIS 23247-1, 2020. [Online]. Available: https://www.iso.org/standard/75066.html[30] Industrial Automation Systems and Integration-Product Data Representation and Exchange—Part 1: Overview and Fundamental Principles, Standard ISO 10303-1, 1994. [Online]. Available: https://www.iso.org/standard/20579.html[31] 2014 Cutting Tool Data Representation and Exchange—Part 3: ReferenceDictionary for Tool Items, Int. Org. Standard, Standard ISO 13399-3,2014. [Online]. Available: https://www.iso.org/standard/54168.html[32] O. Foundation. Unified Architecture. Accessed: 2008. [Online]. Available: https://opcfoundation.org/about/opc-technologies/opc-ua/[33] R. Rosen, G. von Wichert, G. Lo, and K. D. Bettenhausen, ‘‘About theimportance of autonomy and digital twins for the future of manufacturing,’’ IFAC-PapersOnLine, vol. 48, no. 3, pp. 567–572, 2015.[34] J. VachÆlek, L. Bartalský, O. Rovný, D. Ši†mi†ovÆ, M. MorhÆc, andM. Lok†ík, ‘‘The digital twin of an industrial production line within theindustry 4.0 concept,’’ in Proc. 21st Int. Conf. Process Control (PC),Jun. 2017, pp. 258–262.[35] E. Negri, S. Berardi, L. Fumagalli, and M. Macchi, ‘‘MES-integrateddigital twin frameworks,’’ J. Manuf. Syst., vol. 56, pp. 58–71, Jul. 2020.[36] Z. Yin and J. Hou, ‘‘Recent advances on SVM based fault diagnosis andprocess monitoring in complicated industrial processes,’’ Neurocomputing, vol. 174, pp. 643–650, Jan. 2016.[37] L. Bennacer, Y. Amirat, A. Chibani, A. Mellouk, and L. Ciavaglia, ‘‘Selfdiagnosis technique for virtual private networks combining Bayesian networks and case-based reasoning,’’ IEEE Trans. Autom. Sci. Eng., vol. 12,no. 1, pp. 354–366, Jan. 2015.[38] P. Tamilselvan and P. Wang, ‘‘Failure diagnosis using deep belief learning based health state classification,’’ Rel. Eng. Syst. Saf., vol. 115,pp. 124–135, Jul. 2013.[39] Y. Qi, C. Shen, D. Wang, J. Shi, X. Jiang, and Z. Zhu, ‘‘Stacked sparseautoencoder-based deep network for fault diagnosis of rotating machinery,’’ IEEE Access, vol. 5, pp. 15066–15079, 2017.[40] W. Lu, Y. Li, Y. Cheng, D. Meng, B. Liang, and P. Zhou, ‘‘Early faultdetection approach with deep architectures,’’ IEEE Trans. Instrum. Meas.,vol. 67, no. 7, pp. 1679–1689, Jul. 2018.[41] Y. Qi Chen, O. Fink, and G. Sansavini, ‘‘Combined fault location andclassification for power transmission lines fault diagnosis with integrated feature extraction,’’ IEEE Trans. Ind. Electron., vol. 65, no. 1,pp. 561–569, Jan. 2018.[42] H. Darong, K. Lanyan, M. Bo, Z. Ling, and S. Guoxi, ‘‘A new incipientfault diagnosis method combining improved RLS and LMD algorithmfor rolling bearings with strong background noise,’’ IEEE Access, vol. 6,pp. 26001–26010, 2018.[43] Y. Wang, Z. Wei, and J. Yang, ‘‘Feature trend extraction and adaptivedensity peaks search for intelligent fault diagnosis of machines,’’ IEEETrans. Ind. Informat., vol. 15, no. 1, pp. 105–115, Jan. 2019.[44] S. Yin, X. Zhu, and O. Kaynak, ‘‘Improved PLS focused on keyperformance-indicator-related fault diagnosis,’’ IEEE Trans. Ind. Electron., vol. 62, no. 3, pp. 1651–1658, Mar. 2015.[45] Y. Xu, Y. Sun, X. Liu, and Y. Zheng, ‘‘A digital-twin-assisted fault diagnosis using deep transfer learning,’’ IEEE Access, vol. 7, pp. 19990–19999,2019.[46] Y. Wang, R. Xiong, H. Yu, J. Zhang, and Y. Liu, ‘‘Perception of demonstration for automatic programing of robotic assembly: Framework, algorithm, and validation,’’ IEEE/ASME Trans. Mechatronics, vol. 23, no. 3,pp. 1059–1070, Jun. 2018.[47] X. Li, B. He, Y. Zhou, and G. Li, ‘‘Multisource model-driven digital twinsystem of robotic assembly,’’ IEEE Syst. J., early access, Jan. 3, 2020,doi: 10.1109/JSYST.2019.2958874.[48] Y. Fang, C. Peng, P. Lou, Z. Zhou, J. Hu, and J. Yan, ‘‘Digital-twinbased job shop scheduling toward smart manufacturing,’’ IEEE Trans.Ind. Informat., vol. 15, no. 12, pp. 6425–6435, Dec. 2019.[49] S. Scharff. (2019). From Digital Twin to Improved Patient Experience. Accessed: May 8, 2020. [Online]. Available: https://www.siemenshealthineers.com/news/mso-digital-twin-mater.html[50] T. Marchal. (Sep. 2016). VPH: The Ultimate Stage Before Your OwnMedical Digital Twin. Accessed: May 8, 2020. [Online]. Available:https://www.linkedin.com/pulse/vph-ultimate-stage-before-your-ownmedical-digital-twin-marchal/?trk=mp-reader-car[51] C. Copley. (Aug. 2018). Medical Technology Firms Develop ‘Digital Twins’ for Personalized Health Care. Accessed: May 8, 2020.[Online]. Available: https://www.theglobeandmail.com/business/articlemedical-technology-firms-develop-digital-twins-for-personalized/[52] R. Martinez-Velazquez, R. Gamez, and A. El Saddik, ‘‘Cardio twin:A digital twin of the human heart running on the edge,’’ in Proc. IEEEInt. Symp. Med. Meas. Appl. (MeMeA), Jun. 2019, pp. 1–6.[53] J. M. Ospel, G. Gascou, V. Costalat, L. Piergallini, K. A. Blackham,and D. W. Zumofen, ‘‘Comparison of Pipeline embolization device sizingbased on conventional 2D measurements and virtual simulation using theSim&Size software: An agreement study,’’ Amer. J. Neuroradiol., vol. 40,no. 3, pp. 524–530, Feb. 2019.[54] M. Holtmannspotter, M. Martinez-Galdamez, M. Isokangas, R. Ferrara,and M. Sanchez, ‘‘Simulation in clinical practice: First experience withSim&Cure before implantation of flow diverter (pipeline) or web-devicefor the treatment of intracranial aneurysm,’’ in Proc. ABC/WIN, 2017.[55] Y. Feng, J. Zhao, X. Chen, and J. Lin, ‘‘An in silico subject-variabilitystudy of upper airway morphological influence on the airflow regime ina tracheobronchial tree,’’ Bioengineering, vol. 4, no. 4, p. 90, Nov. 2017.[56] Y. Liu, L. Zhang, Y. Yang, L. Zhou, L. Ren, F. Wang, R. Liu, Z. Pang, andM. J. Deen, ‘‘A novel cloud-based framework for the elderly healthcareservices using digital twin,’’ IEEE Access, vol. 7, pp. 49088–49101, 2019.[57] Z. Wang, X. Liao, X. Zhao, K. Han, P. Tiwari, M. J. Barth, and G. Wu,‘‘A digital twin paradigm: Vehicle-to-cloud based advanced driver assistance systems,’’ in Proc. IEEE 91st Veh. Technol. Conf. (VTC-Spring),May 2020, pp. 1–6.[58] E. Cioroaica, T. Kuhn, and B. Buhnova, ‘‘(Do Not) trust in ecosystems,’’in Proc. IEEE/ACM 41st Int. Conf. Softw. Eng., New Ideas Emerg. Results(ICSE-NIER), May 2019, pp. 9–12.[59] X. Chen, T. Chen, Z. Zhao, H. Zhang, M. Bennis, and J. I. Yusheng,‘‘Resource awareness in unmanned aerial vehicle-assisted mobile-edgecomputing systems,’’ in Proc. IEEE 91st Veh. Technol. Conf. (VTCSpring), May 2020, pp. 1–6.[60] S. A. P. Kumar, R. Madhumathi, P. R. Chelliah, L. Tao, and S. Wang,‘‘A novel digital twin-centric approach for driver intention prediction andtraffic congestion avoidance,’’ J. Reliable Intell. Environ., vol. 4, no. 4,pp. 199–209, Dec. 2018.VOLUME 9, 2021 32049M. M. Rathore et al.: Role of AI, Machine Learning, and Big Data in Digital Twinning: A SLR, Challenges, and Opportunities[61] S. Venkatesan, K. Manickavasagam, N. Tengenkai, andN. Vijayalakshmi, ‘‘Health monitoring and prognosis of electricvehicle motor using intelligent-digital twin,’’ IET Electr. Power Appl.,vol. 13, no. 9, pp. 1328–1335, Sep. 2019.[62] S. M. E. Sepasgozar, ‘‘Digital twin and Web-based virtual gaming technologies for online education: A case of construction management andengineering,’’ Appl. Sci., vol. 10, no. 13, p. 4678, Jul. 2020.[63] M. Lammers. (Jun. 2018). Opinion | Digital Twin Offers Huge Opportunities for Real Estate Life Cycle. Accessed: May 8, 2020. [Online].Available: https://www.proptech.nl/blog/digital-twin/[64] A. Kampker, V. Stich, P. Jussen, B. Moser, and J. Kuntz, ‘‘Businessmodels for industrial smart services—The example of a digital twin fora product-service-system for potato harvesting,’’ Procedia CIRP, vol. 83,pp. 534–540, Jan. 2019.[65] S. H. Khajavi, N. H. Motlagh, A. Jaribion, L. C. Werner, andJ. Holmström, ‘‘Digital twin: Vision, benefits, boundaries, and creationfor buildings,’’ IEEE Access, vol. 7, pp. 147406–147419, 2019.[66] C. Fan, C. Zhang, A. Yahja, and A. Mostafavi, ‘‘Disaster city digitaltwin: A vision for integrating artificial and human intelligence for disastermanagement,’’ Int. J. Inf. Manage., vol. 56, Feb. 2021, Art. no. 102049.[67] M. Dietz and G. Pernul, ‘‘Unleashing the digital Twin’s potential for ICSsecurity,’’ IEEE Secur. Privacy, vol. 18, no. 4, pp. 20–27, Jul. 2020.[68] R. Langner, ‘‘To kill a centrifuge: A technical analysis of what stuxnet’screators tried to achieve,’’ The Langner Group, Tech. Rep., 2013.[69] S. Miller, N. Brubaker, D. K. Zafra, and D. Caban, ‘‘Triton actor TTPprofile, custom attack tools, detections, and ATT&CK mapping,’’ FireeyeThreat Res. Blog, Apr. 2019.[70] M. M. U. Rathore, M. J. J. Gul, A. Paul, A. A. Khan, R. W. Ahmad,J. Rodrigues, and S. Bakiras, ‘‘Multilevel graph-based decision making in big scholarly data: An approach to identify expert reviewer,finding quality impact factor, ranking journals and researchers,’’ IEEETrans. Emerg. Topics Comput., early access, Sep. 10, 2018, doi:10.1109/TETC.2018.2869458.[71] M. M. Rathore, H. Son, A. Ahmad, and A. Paul, ‘‘Real-time videoprocessing for traffic control in smart city using Hadoop ecosystem withGPUs,’’ Soft Comput., vol. 22, no. 5, pp. 1533–1544, Mar. 2018.[72] M. M. Rathore, A. Ahmad, A. Paul, and S. Rho, ‘‘Exploiting encryptedand tunneled multimedia calls in high-speed big data environment,’’Multimedia Tools Appl., vol. 77, no. 4, pp. 4959–4984, Feb. 2018.[73] S. A. Shah, D. Z. Seker, M. M. Rathore, S. Hameed, S. Ben Yahia,and D. Draheim, ‘‘Towards disaster resilient smart cities: Can Internetof Things and big data analytics be the game changers?’’ IEEE Access,vol. 7, pp. 91885–91903, 2019.[74] X. Yuan, C. J. Anumba, and M. K. Parfitt, ‘‘Cyber-physical systems fortemporary structure monitoring,’’ Autom. Construct., vol. 66, pp. 1–14,Jun. 2016.[75] F. Thiesse, M. Dierkes, and E. Fleisch, ‘‘LotTrack: RFID-based processcontrol in the semiconductor industry,’’ IEEE Pervas. Comput., vol. 5,no. 1, pp. 47–53, Jan. 2006.[76] H. Choi, Y. Baek, and B. Lee, ‘‘Design and implementation of practicalasset tracking system in container terminals,’’ Int. J. Precis. Eng. Manuf.,vol. 13, no. 11, pp. 1955–1964, Nov. 2012.[77] Y. Zheng, S. Yang, and H. Cheng, ‘‘An application framework of digitaltwin and its case study,’’ J. Ambient Intell. Humanized Comput., vol. 10,no. 3, pp. 1141–1153, Mar. 2019.[78] K. Ding, H. Shi, J. Hui, Y. Liu, B. Zhu, F. Zhang, and W. Cao, ‘‘Smart steelbridge construction enabled by BIM and Internet of Things in industry4.0: A framework,’’ in Proc. IEEE 15th Int. Conf. Netw., Sens. Control(ICNSC), Mar. 2018, pp. 1–5.[79] Y. Zhou, T. Xing, Y. Song, Y. Li, X. Zhu, G. Li, and S. Ding, ‘‘Digitaltwin-driven geometric optimization of centrifugal impeller with free-formblades for five-axis flank milling,’’ J. Manuf. Syst., Jul. 2020.[80] A. Oluwasegun and J.-C. Jung, ‘‘The application of machine learning forthe prognostics and health management of control element drive system,’’Nucl. Eng. Technol., vol. 52, no. 10, pp. 2262–2273, Oct. 2020.[81] A. Gandomi and M. Haider, ‘‘Beyond the hype: Big data concepts,methods, and analytics,’’ Int. J. Inf. Manage., vol. 35, no. 2, pp. 137–144,Apr. 2015.[82] Y. Zhang, S. Ma, H. Yang, J. Lv, and Y. Liu, ‘‘A big data driven analyticalframework for energy-intensive manufacturing industries,’’ J. CleanerProd., vol. 197, pp. 57–72, Oct. 2018.[83] Y. Zhang, S. Ren, Y. Liu, and S. Si, ‘‘A big data analytics architecture forcleaner manufacturing and maintenance processes of complex products,’’J. Cleaner Prod., vol. 142, pp. 626–641, Jan. 2017.[84] Y. Wang, S. Wang, B. Yang, L. Zhu, and F. Liu, ‘‘Big data driven hierarchical digital twin predictive remanufacturing paradigm: Architecture,control mechanism, application scenario and benefits,’’ J. Cleaner Prod.,vol. 248, Mar. 2020, Art. no. 119299.[85] K. M. Alam and A. El Saddik, ‘‘C2PS: A digital twin architecture reference model for the cloud-based cyber-physical systems,’’ IEEE Access,vol. 5, pp. 2050–2062, 2017.[86] E. A. Patterson, R. J. Taylor, and M. Bankhead, ‘‘A framework for anintegrated nuclear digital environment,’’ Prog. Nucl. Energy, vol. 87,pp. 97–103, Mar. 2016.[87] M. Zhang, F. Tao, and A. Y. C. Nee, ‘‘Digital twin enhanced dynamicjob-shop scheduling,’’ J. Manuf. Syst., May 2020.[88] M. Schluse, M. Priggemeyer, L. Atorf, and J. Rossmann, ‘‘Experimentable digital twins—Streamlining simulation-based systems engineering for industry 4.0,’’ IEEE Trans. Ind. Informat., vol. 14, no. 4,pp. 1722–1731, Feb. 2018.[89] S. Zhang, C. Kang, Z. Liu, J. Wu, and C. Ma, ‘‘A product quality monitormodel with the digital twin model and the stacked auto encoder,’’ IEEEAccess, vol. 8, pp. 113826–113836, 2020.[90] R. Bansal, M. A. Khanesar, and D. Branson, ‘‘Ant colony optimizationalgorithm for industrial robot programming in a digital twin,’’ in Proc.25th Int. Conf. Autom. Comput. (ICAC), Sep. 2019, pp. 1–5.[91] K. Xia, C. Sacco, M. Kirkpatrick, C. Saidy, L. Nguyen, A. Kircaliali, andR. Harik, ‘‘A digital twin to train deep reinforcement learning agent forsmart manufacturing plants: Environment, interfaces and intelligence,’’J. Manuf. Syst., Jul. 2020.[92] F. Tao, J. Cheng, Q. Qi, M. Zhang, H. Zhang, and F. Sui, ‘‘Digital twindriven product design, manufacturing and service with big data,’’ Int. J.Adv. Manuf. Technol., vol. 94, nos. 9–12, pp. 3563–3576, Feb. 2018.[93] H. Zhang, G. Zhang, and Q. Yan, ‘‘Digital twin-driven cyber-physical production system towards smart shop-floor,’’ J. Ambient Intell. HumanizedComput., vol. 10, no. 11, pp. 4439–4453, Nov. 2019.[94] W. Wang, Y. Zhang, and R. Y. Zhong, ‘‘A proactive material handlingmethod for CPS enabled shop-floor,’’ Robot. Comput.-Integr. Manuf.,vol. 61, Feb. 2020, Art. no. 101849.[95] C. Zhang, G. Zhou, J. Hu, and J. Li, ‘‘Deep learning-enabled intelligentprocess planning for digital twin manufacturing cell,’’ Knowl.-BasedSyst., vol. 191, Mar. 2020, Art. no. 105247.[96] S. Liu, J. Bao, Y. Lu, J. Li, S. Lu, and X. Sun, ‘‘Digital twin modelingmethod based on biomimicry for machining aerospace components,’’J. Manuf. Syst., May 2020.[97] J. Liu, H. Zhou, G. Tian, X. Liu, and X. Jing, ‘‘Digital twin-based processreuse and evaluation approach for smart process planning,’’ Int. J. Adv.Manuf. Technol., vol. 100, nos. 5–8, pp. 1619–1634, Feb. 2019.[98] P. Franciosa, M. Sokolov, S. Sinha, T. Sun, and D. Ceglarek, ‘‘Deeplearning enhanced digital twin for remote laser welding of aluminiumstructures,’’ CIRP Ann. Manuf. Technol., vol. 69, no. 1, 2020.[99] Y. Golizadeh Akhlaghi, A. Badiei, X. Zhao, K. Aslansefat, X. Xiao,S. Shittu, and X. Ma, ‘‘A constraint multi-objective evolutionary optimization of a state-of-the-art dew point cooler using digital twins,’’Energy Convers. Manage., vol. 211, May 2020, Art. no. 112772.[100] C. Liu, L. Le Roux, C. Körner, O. Tabaste, F. Lacan, and S. Bigot,‘‘Digital twin-enabled collaborative data management for metal additivemanufacturing systems,’’ J. Manuf. Syst., May 2020.[101] V. Mnih, K. Kavukcuoglu, D. Silver, A. Graves, I. Antonoglou,D. Wierstra, and M. Riedmiller, ‘‘Playing atari with deep reinforcementlearning,’’ 2013, arXiv:1312.5602. [Online]. Available: http://arxiv.org/abs/1312.5602[102] H. Van Hasselt, A. Guez, and D. Silver, ‘‘Deep reinforcement learningwith double q-learning,’’ in Proc. 13th AAAI Conf. Artif. Intell., 2016,pp. 1–7.[103] T. Schaul, J. Quan, I. Antonoglou, and D. Silver, ‘‘Prioritized experiencereplay,’’ 2015, arXiv:1511.05952. [Online]. Available: http://arxiv.org/abs/1511.05952[104] J. Leng, C. Jin, A. Vogl, and H. Liu, ‘‘Deep reinforcement learningfor a color-batching resequencing problem,’’ J. Manuf. Syst., vol. 56,pp. 175–187, Jul. 2020.[105] K. He, X. Zhang, S. Ren, and J. Sun, ‘‘Identity mappings in deep residualnetworks,’’ in Proc. Eur. Conf. Comput. Vis. Cham, Switzerland: Springer,2016, pp. 630–645.[106] J. Lee, M. Azamfar, J. Singh, and S. Siahpour, ‘‘Integration of digitaltwin and deep learning in cyber-physical systems: Towards smart manufacturing,’’ IET Collaborative Intell. Manuf., vol. 2, no. 1, pp. 34–36,Mar. 2020.32050 VOLUME 9, 2021M. M. Rathore et al.: Role of AI, Machine Learning, and Big Data in Digital Twinning: A SLR, Challenges, and Opportunities[107] M. Tomko and S. Winter, ‘‘Beyond digital twins—A commentary,’’ Environ. Planning B, Urban Analytics City Sci., vol. 46, no. 2, pp. 395–399,Feb. 2019.[108] F. Tao, M. Zhang, Y. Liu, and A. Y. C. Nee, ‘‘Digital twin driven prognostics and health management for complex equipment,’’ CIRP Ann., vol. 67,no. 1, pp. 169–172, 2018.[109] G.-B. Huang, H. Zhou, X. Ding, and R. Zhang, ‘‘Extreme learningmachine for regression and multiclass classification,’’ IEEE Trans. Syst.,Man, Cybern. B, Cybern., vol. 42, no. 2, pp. 513–529, Apr. 2012.[110] A. Coraddu, L. Oneto, F. Baldi, F. Cipollini, M. Atlar, and S. Savio,‘‘Data-driven ship digital twin for estimating the speed loss caused bythe marine fouling,’’ Ocean Eng., vol. 186, Aug. 2019, Art. no. 106063.[111] J. Tang, C. Deng, and G.-B. Huang, ‘‘Extreme learning machine formultilayer perceptron,’’ IEEE Trans. Neural Netw. Learn. Syst., vol. 27,no. 4, pp. 809–821, Apr. 2016.[112] P. Jain, J. Poon, J. P. Singh, C. Spanos, S. R. Sanders, and S. K. Panda,‘‘A digital twin approach for fault diagnosis in distributed photovoltaicsystems,’’ IEEE Trans. Power Electron., vol. 35, no. 1, pp. 940–956,Jan. 2020.[113] W. Li, M. Rentemeister, J. Badeda, D. Jöst, D. Schulte, and D. U. Sauer,‘‘Digital twin for battery systems: Cloud battery management system withonline state-of-charge and state-of-health estimation,’’ J. Energy Storage,vol. 30, Aug. 2020, Art. no. 101557.[114] A. Piros, L. Trautmann, and E. Baka, ‘‘Error handling method for digitaltwin-based plasma radiation detection,’’ Fusion Eng. Design, vol. 156,Jul. 2020, Art. no. 111592.[115] M. G. Kapteyn, D. J. Knezevic, D. B. P. Huynh, M. Tran, andK. E. Willcox, ‘‘Data-driven physics-based digital twins via a library ofcomponent-based reduced-order models,’’ Int. J. Numer. Methods Eng.,Jun. 2020.[116] Y. Ye, Q. Yang, F. Yang, Y. Huo, and S. Meng, ‘‘Digital twin for thestructural health management of reusable spacecraft: A case study,’’ Eng.Fract. Mech., vol. 234, Jul. 2020, Art. no. 107076.[117] K. P. Murphy, ‘‘Dynamic Bayesian networks: Representation, inferenceand learning, dissertation,’’ Ph.D. dissertation, Dept. Comput. Sci., Univ.California, Berkeley Berkeley, CA, USA, 2002.[118] P. E. Leser, J. E. Warner, W. P. Leser, G. F. Bomarito, J. A. Newman,and J. D. Hochhalter, ‘‘A digital twin feasibility study (Part II): Nondeterministic predictions of fatigue life using in-situ diagnostics andprognostics,’’ Eng. Fract. Mech., vol. 229, Apr. 2020, Art. no. 106903.[119] H. Zhang, Q. Yan, and Z. Wen, ‘‘Information modeling for cyber-physicalproduction system based on digital twin and automationml,’’ Int. J. Adv.Manuf. Technol., pp. 1–19, Mar. 2020.[120] Z. Liu, W. Chen, C. Zhang, C. Yang, and H. Chu, ‘‘Data super-networkfault prediction model and maintenance strategy for mechanical productbased on digital twin,’’ IEEE Access, vol. 7, pp. 177284–177296, 2019.[121] W. Booyse, D. N. Wilke, and S. Heyns, ‘‘Deep digital twins for detection,diagnostics and prognostics,’’ Mech. Syst. Signal Process., vol. 140,Jun. 2020, Art. no. 106612.[122] H. Kim, C. Jin, M. Kim, and K. Kim, ‘‘Damage detection of bottom-setgillnet using artificial neural network,’’ Ocean Eng., vol. 208, Jul. 2020,Art. no. 107423.[123] W. Luo, T. Hu, C. Zhang, and Y. Wei, ‘‘Digital twin for CNC machinetool: Modeling and using strategy,’’ J. Ambient Intell. Humanized Comput., vol. 10, no. 3, pp. 1129–1140, Mar. 2019.[124] W. Luo, T. Hu, Y. Ye, C. Zhang, and Y. Wei, ‘‘A hybrid predictivemaintenance approach for CNC machine tool driven by digital twin,’’Robot. Comput.-Integr. Manuf., vol. 65, Oct. 2020, Art. no. 101974.[125] X. Song, T. Jiang, S. Schlegel, and D. Westermann, ‘‘Parameter tuningfor dynamic digital twins in inverter-dominated distribution grid,’’ IETRenew. Power Gener., vol. 14, no. 5, pp. 811–821, Apr. 2020.[126] S. K. Andryushkevich, S. P. Kovalyov, and E. Nefedov, ‘‘Compositionand application of power system digital twins based on ontological modeling,’’ in Proc. IEEE 17th Int. Conf. Ind. Informat. (INDIN), vol. 1,Jul. 2019, pp. 1536–1542.[127] D. Gong, J. Sun, and Z. Miao, ‘‘A set-based genetic algorithm for intervalmany-objective optimization problems,’’ IEEE Trans. Evol. Comput.,vol. 22, no. 1, pp. 47–60, Feb. 2018.[128] M. Zhou, J. Yan, and D. Feng, ‘‘Digital twin framework and its applicationto power grid online analysis,’’ CSEE J. Power Energy Syst., vol. 5, no. 3,pp. 391–398, 2019.[129] C. C. Lee, ‘‘Fuzzy logic in control systems: Fuzzy logic controller. II,’’IEEE Trans. Syst., Man, Cybern., vol. 20, no. 2, pp. 419–435, Apr. 1990.[130] J. Morton, T. A. Wheeler, and M. J. Kochenderfer, ‘‘Analysis of recurrentneural networks for probabilistic modeling of driver behavior,’’ IEEETrans. Intell. Transp. Syst., vol. 18, no. 5, pp. 1289–1298, May 2017.[131] X. Chen, C. Wu, T. Chen, H. Zhang, Z. Liu, Y. Zhang, and M. Bennis,‘‘Age of information aware radio resource management in vehicularnetworks: A proactive deep reinforcement learning perspective,’’ IEEETrans. Wireless Commun., vol. 19, no. 4, pp. 2268–2281, Apr. 2020.[132] L. Zhao, G. Han, Z. Li, and L. Shu, ‘‘Intelligent digital twin-basedsoftware-defined vehicular networks,’’ IEEE Netw., vol. 34, no. 5,pp. 178–184, Sep. 2020.[133] B. R. Barricelli, E. Casiraghi, J. Gliozzo, A. Petrini, and S. Valtolina,‘‘Human digital twin for fitness management,’’ IEEE Access, vol. 8,pp. 26637–26664, 2020.[134] O. Mazumder, D. Roy, S. Bhattacharya, A. Sinha, and A. Pal, ‘‘SyntheticPPG generation from haemodynamic model with baroreflex autoregulation: A digital twin of cardiovascular system,’’ in Proc. 41st Annu. Int.Conf. IEEE Eng. Med. Biol. Soc. (EMBC), Jul. 2019, pp. 5024–5029.[135] N. K. Chakshu, J. Carson, I. Sazonov, and P. Nithiarasu, ‘‘A semiactive human digital twin model for detecting severity of carotid stenosesfrom head vibration—A coupled computational mechanics and computervision method,’’ Int. J. Numer. methods Biomed. Eng., vol. 35, no. 5,p. e3180, 2019.[136] F. Laamarti, H. Faiz Badawi, Y. Ding, F. Arafsha, B. Hafidh, andA. El Saddik, ‘‘An ISO/IEEE 11073 standardized digital twin framework for health and well-being in smart cities,’’ IEEE Access, vol. 8,pp. 105950–105961, 2020.[137] N. S. Altman, ‘‘An introduction to kernel and nearest-neighbor nonparametric regression,’’ Amer. Statistician, vol. 46, no. 3, pp. 175–185,Aug. 1992.[138] C. Cortes and V. Vapnik, ‘‘Support-vector networks,’’ Mach. Learn.,vol. 20, no. 3, pp. 273–297, 1995.[139] M. Pengnoo, M. Taynnan Barros, L. Wuttisittikulkij, B. Butler, A. Davy,and S. Balasubramaniam, ‘‘Digital twin for metasurface reflector management in 6G terahertz communications,’’ IEEE Access, vol. 8,pp. 114580–114596, 2020.[140] R. Zhang, Y. Yang, W. Wang, L. Zeng, J. Chen, and S. McGrath, ‘‘An algorithm for obstacle detection based on YOLO and light filed camera,’’ inProc. 12th Int. Conf. Sens. Technol. (ICST), Dec. 2018, pp. 223–226.[141] G. Schrotter and C. Hürzeler, ‘‘The digital twin of the city of Zurichfor urban planning,’’ PFG, J. Photogramm., Remote Sens. Geoinf. Sci.,pp. 1–14, Feb. 2020.[142] H. Lehner and L. Dorffner, ‘‘Digital geoTwin Vienna: Towards a digitaltwin city as Geodata Hub,’’ PFG, J. Photogramm., Remote SensingGeoinformat. Sci. volume, vol. 88, pp. 63–75, 2020.[143] J. Döllner, ‘‘Geospatial artificial intelligence: Potentials of machine learning for 3D point clouds and geospatial digital twins,’’ PFG, J. Photogramm., Remote Sens. Geoinformation Sci., pp. 1–10, 2020.[144] X. Tong, Q. Liu, S. Pi, and Y. Xiao, ‘‘Real-time machining data application and service based on IMT digital twin,’’ J. Intell. Manuf., vol. 8,pp. 1–20, Oct. 2019.[145] A. M. Lund, K. Mochel, J. Lin, R. Onetto, J. Srinivasan, P. Gregg,J. E. Bergman, K. D. Hartling, A. Ahmed, and S. Chotai, ‘‘Digital systemand method for managing a wind farm having plurality of wind turbinescoupled to power grid,’’ U.S. Patent 10 132 295, Nov. 20, 2018.[146] T. Shah, S. Govindappa, P. Nistler, and B. Narayanan, ‘‘Digital twinsystem for a cooling system,’’ U.S. Patent 9 881 430, Jan. 30, 2018.[147] H. Wang, ‘‘Digital twin based management system and method anddigital twin based fuel cell management system and method,’’ U.S. Patent10 522 854, Dec. 31, 2019.[148] J. E. Hershey, F. W. Wheeler, M. C. Nielsen, C. D. Johnson,M. J. Dell’Anno, and J. Joykutti, ‘‘Digital twin of twinned physical system,’’ U.S. Patent App. 15 087 217, Oct. 5, 2017.[149] Z. Song and A. M. Canedo, ‘‘Digital twins for energy efficient assetmaintenance,’’ U.S. Patent App. 15 052 992, Aug. 25, 2016.[150] C. J. Yates, M. Stankiewicz, J. Alexander, and C. Softley, ‘‘Industrialsafety monitoring configuration using a digital twin,’’ U.S. Patent App.16 189 116, May 14, 2020.[151] T. Masuda, B. Kim, and S. Shiraishi, ‘‘Proactive vehicle maintenance scheduling based on digital twin simulations,’’ U.S. Patent App.15 908 768, Aug. 29, 2019.[152] H. Goldfarb, A. Pandey, and W. Yan, ‘‘Feature selection and featuresynthesis methods for predictive modeling in a twinned physical system,’’U.S. Patent App. 15 350 665, May 17, 2018.VOLUME 9, 2021 32051M. M. Rathore et al.: Role of AI, Machine Learning, and Big Data in Digital Twinning: A SLR, Challenges, and Opportunities[153] J. Zimmerman, C. Dodd, and M. Peterson, ‘‘Methods and systems for generating a patient digital twin,’’ U.S. Patent App. 15 635 805, Jan. 3, 2019.[154] S. Nagesh, ‘‘X-ray tube bearing failure prediction using digital twinanalytics,’’ U.S. Patent 10 524 760, Jan. 7, 2020.[155] M. Peterson, ‘‘Surgery digital twin,’’ U.S. Patent App. 15 711 786,Mar. 21, 2019.[156] L. G. E. Cox, C. P. Hendriks, M. Bulut, V. Lavezzo, and O. van der Sluis,‘‘Digital twin operation,’’ U.S. Patent App. 16 704 495, Jun. 11, 2020.[157] K. Fischer and M. Heintel, ‘‘Examining a consistency between referencedata of a production object and data of a digital twin of the productionobject,’’ U.S. Patent App. 15 750 538, Aug. 9, 2018.[158] M. G. Burd and P. F. McLaughlin, ‘‘Integrated digital twin for an industrial facility,’’ U.S. Patent App. 15 416 569, Jul. 26, 2018.[159] K. Deutsch, S. Pal, R. Milev, and K. Yang, ‘‘Contextual digital twinruntime environment,’’ U.S. Patent 10 564 993, Feb. 18, 2020.[160] S. Shiraishi, Z. Jiang, and B. Kim, ‘‘Digital twin for vehicle risk evaluation,’’ U.S. Patent App. 16 007 693, Dec. 19, 2019.[161] S. Shiraishi and Y. Zhao, ‘‘Sensor-based digital twin system for vehicularanalysis,’’ U.S. Patent App. 15 925 070, Sep. 19, 2019.[162] A. Yousif, A. Ayyagari, D. T. Kirkland, E. C. Owyang, J. Apanovitch,and T. W. Anstey, ‘‘Aircraft communications system with an operationaldigital twin,’’ U.S. Patent App. 16 100 985, Feb. 13, 2020.[163] Y. Park, S. R. Sinha, V. Venkiteswaran, V. S. Chennupati, andE. S. Paulson, ‘‘Building system with digital twin based data ingestionand processing,’’ U.S. Patent 10 854 194, Dec. 1, 2020.[164] Q. Min, Y. Lu, Z. Liu, C. Su, and B. Wang, ‘‘Machine learning baseddigital twin framework for production optimization in petrochemicalindustry,’’ Int. J. Inf. Manage., vol. 49, pp. 502–519, Dec. 2019.[165] S. Bangsow, Tecnomatix Plant Simulation. Springer, 2015.[166] A. Glikson, ‘‘Fi-Ware: Core platform for future Internet applications,’’ inProc. 4th Annu. Int. Conf. Syst. Storage, 2011.[167] A. Bosch Rexroth. Indramotion Mtx. Accessed: 2010.https://www.boschrexroth.com/en/us/products/product-groups/electricdrives-and-controls/topics/cnc/indramotion-mtx-standard-performanceand-advanced/index[168] T. White, Hadoop: The Definitive Guide. Newton, MA, USA:O’Reilly Media, 2012.[169] M. Abadi, P. Barham, J. Chen, Z. Chen, A. Davis, J. Dean, M. Devin,S. Ghemawat, G. Irving, M. Isard, and M. Kudlur, ‘‘TensorFlow: A system for large-scale machine learning,’’ in Proc. 12th USENIX Symp.Operating Syst. Design Implement. (OSDI), 2016, pp. 265–283.[170] F. Seide and A. Agarwal, ‘‘CNTK: Microsoft’s open-source deep-learningtoolkit,’’ in Proc. 22nd ACM SIGKDD Int. Conf. Knowl. Discovery DataMining, Aug. 2016, p. 2135.[171] Y. Jia, E. Shelhamer, J. Donahue, S. Karayev, J. Long, R. Girshick,S. Guadarrama, and T. Darrell, ‘‘Caffe: Convolutional architecture forfast feature embedding,’’ in Proc. 22nd ACM Int. Conf. Multimedia,Nov. 2014, pp. 675–678.[172] A. Gulli and S. Pal, Deep Learning With Keras. Birmingham, U.K.: Packt,2017.[173] M. Hall, E. Frank, G. Holmes, B. Pfahringer, P. Reutemann, andI. H. Witten, ‘‘The WEKA data mining software: An update,’’ ACMSIGKDD Explor. Newslett., vol. 11, no. 1, pp. 10–18, 2009.[174] G. Brockman, V. Cheung, L. Pettersson, J. Schneider, J. Schulman,J. Tang, and W. Zaremba, ‘‘OpenAI gym,’’ 2016, arXiv:1606.01540.[Online]. Available: http://arxiv.org/abs/1606.01540[175] Y. Duan, X. Chen, R. Houthooft, J. Schulman, and P. Abbeel, ‘‘Benchmarking deep reinforcement learning for continuous control,’’ in Proc.Int. Conf. Mach. Learn., 2016, pp. 1329–1338.M. MAZHAR RATHORE (Member, IEEE)received the master’s degree in computer and communication security from the National Universityof Sciences and Technology, Pakistan, in 2012,and the Ph.D. degree in computer science andengineering from Kyungpook National University,South Korea, in 2018. He is currently workingas a Postdoctoral Researcher with the College ofScience and Engineering, Hamad Bin Khalifa University, Qatar. His research interests include bigdata analytics, the Internet of Things, smart systems, network traffic analysisand monitoring, remote sensing, smart city, urban planning, intrusiondetection, and information security and privacy. He is a professional memberof ACM. He received the Best Project/Paper Award in the 2016 Qualcomm Innovation Award at Kyungpook National University, for his paper‘‘IoT-Based Smart City Development Using Big Data Analytical Approach.’’He was also a nominee for the Best Project Award in the 2015 IEEECommunications Society Student Competition, for his Project ‘‘IoT-BasedSmart City.’’ He is serving frequently as a Reviewer for various IEEE, ACM,Springer, and Elsevier journals.SYED ATTIQUE SHAH received the Ph.D. degreefrom the Institute of Informatics, Istanbul Technical University, Istanbul, Turkey. During his Ph.D.degree, he was a Visiting Scholar with NationalChiao Tung University, Taiwan, The University ofTokyo, Japan, and the Tallinn University of Technology, Estonia, where he completed the majorcontent of his thesis. He was an Assistant Professorand the Chairperson of the Department of Computer Science, BUITEMS, Quetta, Pakistan. He iscurrently a Lecturer with the Data Systems Group, Institute of ComputerScience, University of Tartu, Estonia. His research interests include big dataanalytics, cloud computing, information management, and the Internet ofThings.DHIRENDRA SHUKLA is currently a Professorand the Dr. J Herbert Smith ACOA Chair in technology management and entrepreneurship of theUniversity of New Brunswick (UNB), Canada.He utilizes his expertise from the telecom sector and extensive academic background in theareas of entrepreneurial finance, masters of business administration, and engineering, to promotea bright future for New Brunswick. Recognition of his tireless efforts and vision are demonstrated through the UNB’s 2014 Award from Startup Canada as the ‘‘MostEntrepreneurial Post-Secondary Institution of the Year,’’ his nomination as aFinalist for the Industry Champion by KIRA, and his nomination as a Finalistfor the Progress Media’s Innovation in Practice Award. He was nominated forthe RBC Top 25 Canadian Immigrant Award and selected by a panel of judgesas a Top 75 finalist. Most recently, he received the Entrepreneur PromotionAward by Startup Canada in 2017, as well as the Outstanding EducatorAward by the Association of Professional Engineers and Geoscientists ofNew Brunswick in 2018.ELMAHDI BENTAFAT received the bachelor’sand M.Sc. degrees in computer science from theEcole Nationale SupØrieure d’Informatique, Algeria, in 2012 and 2016, respectively. He is currentlypursuing the Ph.D. degree with the College of Science and Engineering, Hamad Bin Khalifa University, Qatar. His research interests include appliedcryptography, privacy, information security, andnetwork security.SPIRIDON BAKIRAS (Member, IEEE) receivedthe B.S. degree in electrical and computer engineering from the National Technical University ofAthens, in 1993, the M.S. degree in telematicsfrom the University of Surrey, in 1994, and thePh.D. degree in electrical engineering from theUniversity of Southern California, in 2000. He iscurrently an Associate Professor with the Collegeof Science and Engineering, Hamad Bin KhalifaUniversity, Qatar. Before that, he held teachingand research positions at Michigan Technological University, The City University of New York, The University of Hong Kong, and The Hong KongUniversity of Science and Technology. His current research interests includesecurity and privacy, applied cryptography, mobile computing, and spatiotemporal databases. He is a member of ACM. He was a recipient of theU.S. National Science Foundation (NSF) CAREER Award.32052 VOLUME 9, 2021
