The Impact of Industry 4.0 Technologies on Key Performance Indicators for a Resilient Supply Chain 4.0

: The term “Resilient Supply Chain 4.0” incorporates two research areas: Industry 4.0 and Supply Chain Resilience (SCRes). Industry 4.0 technologies include innovations such as the Internet of Things (IoT), Cyber-Physical Systems (CPS), Augmented Reality (AR), Cloud Computing (CC), the Internet of Services (IoS), Big Data Analytics (BDA), Artiﬁcial Intelligence (AI), Digital Twins (DT), Blockchain (BC), Industrial Robotics (IR), and Additive Manufacturing (AM). Industry 4.0 technologies do not have a direct impact on SCRes, but on resilience elements such as ﬂexibility, redundancy, visibility, agility, collaboration, robustness, and information sharing. This paper aims to investigate which of the Industry 4.0 technologies can help improve the Key Performance Indicators (KPIs) that are used for creating a Resilient Supply Chain 4.0. A non-systematic literature review has been conducted for the identiﬁcation of (a) the most important constituent elements of SCRes, (b) the Industry 4.0 technologies that improve the SCRes elements, and (c) the KPIs that enhance SCRes. A systematic literature review has been conducted to identify which of the Industry 4.0 technologies have an impact on the KPIs that enhance SCRes. The ﬁndings of this work demonstrate that Industry 4.0 technologies can help improve the KPIs for a Resilient Supply Chain 4.0.


Introduction
Major disruptions with a frequent occurrence expose supply chains (SCs) to changes, increasing their vulnerability. The categories of risks that supply chains are facing are [1,2]: (a) external or environmental risks (e.g., pandemics, geopolitical conflicts, natural disasters, intentional attacks, and cyberattacks), which are unexpected and unavoidable and organizations can only be prepared to mitigate the disruptions, and (b) internal risks, which include production risks (e.g., equipment malfunctions), management risks (e.g., lack of skilled staff, coordination problems, productivity problems), information risks (e.g., loss of data, information system failure), and supply and demand risks (e.g., customer change orders, delays in delivery). Resilience capabilities in SCs enable adaptation to the disruptions and restoration of normal supply network operations [3,4].
Industry 4.0 expresses the fourth industrial revolution that characterizes the 21st century. The term "Industry 4.0" was introduced in Germany in 2011 by a Working Group of academics and industry professionals with the aim of studying the impact of innovative Digital Technology systems on manufacturing [5]. Examples of core Industry 4.0 technologies are the Internet of Things (IoT), Cyber-Physical Systems (CPS), Big Data To answer RQ1, first we need to investigate which are the most important SCRes constituent elements, and then how each of the selected Industry 4.0 technologies influence each of these SCRes elements. Additionally, to answer RQ2, first we need to search for publications that identify KPIs for SCRes, focusing on the relation between SCRes constituent elements and KPIs. We then need to exploit the results of RQ1 to determine how each of the selected Industry 4.0 technologies affect the selected KPIs.
In the following, we describe in detail the steps of the literature review that was conducted in this study. The gradual implementation of the overall research methodology is graphically illustrated in Figure 1. To answer RQ1, first we need to investigate which are the most important SCRes constituent elements, and then how each of the selected Industry 4.0 technologies influence each of these SCRes elements. Additionally, to answer RQ2, first we need to search for publications that identify KPIs for SCRes, focusing on the relation between SCRes constituent elements and KPIs. We then need to exploit the results of RQ1 to determine how each of the selected Industry 4.0 technologies affect the selected KPIs.
In the following, we describe in detail the steps of the literature review that was conducted in this study. The gradual implementation of the overall research methodology is graphically illustrated in Figure 1. We started non-systematically searching Google Scholar and Scopus databases to find some recently published papers that discuss the origins of SCRes [2,3,120], and we focused on publications that identify the constituent elements of SCRes [2,3,. During this investigation, we selected some definitions of SCRes [3,[125][126][127]134,142]. Section 3 presents the results of this step of our investigation.
Regarding RQ2, we searched systematically for publications that identify the impact of Industry 4.0 technologies on KPIs used for creating a resilient SC. The keywords were not determined from the beginning but emerged during the research process of the previous steps and particularly after answering RQ1. More specifically, we searched Scopus database using the following search rule: TITLE-ABS-KEY (("supply chain" OR "supply chain 4.0" OR "smart supply chain" OR "digital supply chain") AND (resilien* OR risk) AND ("industry 4.0" OR "digital industry" OR *4.0 OR smart* OR IoT OR "Internet of Things" OR IoS OR "Internet of Services" OR "Cyber-Physical*" OR "Cyber Physical*" OR CPS OR "Artificial Intelligence" OR "Big Data" OR "Augmented Reality" OR "3D print*" OR "Additive Manufactur*" OR "Digital Twin*" OR blockchain* OR "cloud computing" OR robot*) AND ((performance AND (indicator* OR metric* OR measure*)) OR kpi*)).
Most relevant to RQ2 are only 5 out of 13 publications [92,117,146,149,150]. Dev et al. [92] propose the monitoring of KPIs to help create SCRes in a smart factory framework consisting of Industry 4.0 technologies. KPIs that are related with AM technology are considered in reverse logistics chain scenarios. The study conducted by Catellani and Bottani [149] identified KPIs for Lean, Agile, Resilient, and Green SC, and evaluated them. The role of Industry 4.0 technologies in SC performance was investigated in parallel but was not connected to each of the evaluated KPIs. Dumitrascu et al. [150] used AI techniques to predict which KPI should be linked to an identified problem within the organization. Monitoring KPIs, the trend line can be observed, and potential risks can be eliminated, increasing the SCRes. Ansari and Kohl [146] provided an approach based on AI for the analysis of KPIs for SCRes and manufacturing resilience. A recent study conducted by Patidar et al. [117] identified KPIs of a resilient SC and evaluated them under the perspective of Industry 4.0 technologies and sustainability. We have adopted the list of KPIs proposed in [117] and we have attempted further research aiming to determine which of the Industry 4.0 technologies affect these KPIs. Table 1 includes the aforementioned publications that identify KPIs for building a Resilient Supply Chain 4.0, and their main research objectives. The differentiation of our study is also mentioned.
Therefore, this study is built upon the work of Patidar et al. [117] to extend current results on how the KPIs for SCRes are influenced by the most important Industry 4.0 technologies. To achieve this objective, we first searched which of the Industry 4.0 technologies have a direct impact on the identified KPIs. We then explored the relation between the SCRes elements and the KPIs, and combined the results of RQ1 to give the Industry 4.0 technologies that improve each of the SCRes elements and indirectly improve the KPIs. Section 5 presents the results of the investigation related with the RQ2.

Definitions of Supply Chain Resilience
The concept of organization resilience [121,131,132] was determined earlier than that of SCRes. Mallak in 1998 [132] stated: "to be resilient (an individual or an organization) need to learn how to quickly design and implement positive adaptive behaviors matched to the immediate situation, while enduring minimal stress". According to Mallak [132], positive adaptive behavior perceives change as an opportunity, adapting responses to the needs of the situation.
Ali et al. [120] noticed that the foundation of the SCRes literature is mainly based on the findings reported in [3,125,135]. In 2003, Rice and Caniato [3] introduced the term "resilient supply network", stating: "A supply network is resilient when it can respond to unexpected disruptions and restore normal supply network operations". The term Supply Chain Resilience (SCRes) is mostly preferred in the literature.
Several other definitions of SCRes have been stated by researchers. We have selected six definitions of the term SCRes from 2003 to 2022, as shown in Table 2. Christopher and Peck, in 2004 [125], gave a brief but comprehensive definition, defining SCRes as the ability of a system to return to its original state or move to a new, more desirable state after being disturbed. In 2009, Ponomarov and Holocomb [134] gave a more descriptive definition, adding the ability of the supply chain to be prepared for unexpected events, except for the ability to respond to disruptions and recover from them. In 2012, Ponis and Koronis [142] extended previous definitions, stating that when a resilient SC network anticipates unexpected disruption, it gains a competitive advantage. Hohenstein et al., in 2015 [127], also added that a resilient SC returns to its original situation or grows by moving to a new, more desirable state, with the objective of increasing customer service, market share, and financial performance. Another interesting definition was given by Datta, in 2017 [126], with SCRes as a dynamic process of directing activities that protect the organization from disruptions, and in case a disruption occurs, trigger a very rapid response that maintains or regains a dynamically stable state, thereby enabling the adaptation of operations to the new requirements before the competitors.

Author(s) Definition of Supply Chain Resilience
Rice and Caniato, 2003 [3] "A supply network is resilient when it can respond to unexpected disruptions and restore normal supply network operations".
Christopher and Peck, 2004 [125] "The ability of a system to return to its original state or move to a new, more desirable state after being disturbed".
Ponomarov and Holocomb, 2009 [134] "The adaptive capability of the supply chain to prepare for unexpected events, respond to disruptions, and recover from them by maintaining continuity of operations at the desired level of connectedness and control over structure and function".
Ponis and Koronis, 2012 [142] "The ability to proactively plan and design the Supply Chain network for anticipating unexpected disruptive (negative) events, respond adaptively to disruptions while maintaining control over structure and function and transcending to a post-event robust state of operations, if possible, more favorable than the one prior to the event, thus gaining competitive advantage".
Hohenstein et al., 2015 [127] "Supply chain's ability to be prepared for unexpected risk events, responding and recovering quickly to potential disruptions to return to its original situation or grow by moving to a new, more desirable state in order to increase customer service, market share and financial performance".
Datta, 2017 [126] "Supply chain resilience is a dynamic process of steering the actions so that the organisation always stays out of danger zone, and if the disruptive/uncertain event occurs, resilience implies initiating a very rapid and efficient response to minimise the consequences and maintaining or regaining a dynamically stable state, which allows it to adapt operations to the requirements of the changed environment before the competitors and succeed in the long run".

Capabilities and Elements of Supply Chain Resilience
In the literature, we can find different lists of elements that enhance SCRes. Researchers agree that Industry 4.0 technologies influence SCRes indirectly, affecting specific elements [96]. For this reason, it is crucial to identify these elements first, and then determine which of them are improved by specific Industry 4.0 technologies. The elements are also referred as indicators [160,161], enablers [133,136], enhancers [137], capabilities [138], or antecedents [96]. Many researchers have contributed, adding more elements, grouping elements into levels, or prioritizing the most important elements.
Ponis and Koronis in 2012 [142] propose that the most important elements of SCRes are agility, flexibility, velocity, visibility, availability, redundancy, mobilization of resources, collaboration, and SC structure knowledge. The authors proposed a grouping of these elements in four structural elements, namely, agility, redundancy, collaboration, and knowledge of SC structure. They also clarified that agility includes flexibility and velocity, redundancy is differentiated from flexibility and agility, and collaboration includes information sharing, collaborative work, joint decision making, and visibility. According to Scholten and Schilder [143], collaboration in the form of information sharing can increase SCRes via increased visibility, velocity, and flexibility.
Ali et al. in 2017 [120] analyzed 103 articles from 2003 to 2015 and concluded that the definition of SCRes is three dimensional. The first dimension is related to the time of reaction: before, during, or after disruption. The second dimension concerns the strategies that are applied--namely, proactive, concurrent, and reactive. The proactive strategy concerns the pre-disruption phase, the concurrent strategy concerns the phase during the disruption, and the reactive strategy concerns the post-disruption phase. The third dimension is related with the capabilities that are necessary for a resilient supply chain. The capabilities include one proactive capability, i.e., the ability to anticipate risk; two concurrent capabilities, i.e., the ability to adapt and the ability to respond; and two reactive capabilities, i.e., the ability to recover and the ability to learn. The authors [120] also identified 27 elements that build SCRes capabilities. Out of these 27 elements, 13 essential elements are selected to support the five resilient capabilities. The ability to anticipate is supported by situation awareness, robustness, increasing visibility, and knowledge management (pre-disruption). The ability to adapt is supported by increasing flexibility and building redundancy. The ability to respond is supported by collaboration and agility. The ability to recover is supported by contingency planning and market position. Finally, the ability to learn is supported by knowledge management (post-disruption) and building social capital.
In 2009, Ponomarov and Holocomb [134] studied the concept of SCRes as a multidimensional phenomenon. Social, psychological, and economic dimensions were explored. Three phases of SCRes were determined: readiness, response, and recovery. Three psychological principles were determined: control, coherence, and connectedness. Control is the guidance and adjustment of strategic and tactical actions throughout the supply network. Coherence is the deeper understanding of a disruption or a threat. Connectedness is the tendency of people to gather when a crisis occurs. A set of elements are then grouped into these phases and categories. For example, flexibility, agility, and risk-sharing are grouped into the response phase and coherence principle.
Numerous other studies propose lists of elements for SCRes and the grouping of these elements. For example, we could mention the studies that are reported by Christopher [161], and Spieske and Birkel in 2021 [96]. Table 3 shows the most essential elements of SCRes and their description according to the literature.

Resilient Supply Chain 4.0
Legner et al. [22] clarified the difference between the terms digitization and digitalization. Digitization is the process associated with converting analogue signals into a digital form, while digitalization refers to "the sociotechnical phenomena and processes of adopting of these (digital) technologies in broader individual, organizational and societal perspectives". Queiroz et al. [53] argue that digitalization is a superset of digitization. In the context of SC management, digitation refers to the adoption of capabilities concerning digital technologies, while in digitalization these adopted capabilities impact all stakeholders of the SC and can enable the improvement of an organization's competitiveness.
In the following, we will investigate the impact of these selected Industry 4.0 technologies on SCRes based on previous literature. The reader can find a long list of Industry 4.0 technologies in Mittal et al. in 2019 [19], in Amoozad Mahdiraji et al. in 2022 [157], and in published reviews on Industry 4.0 technologies, such as the reviews reported by Lu in 2017 [16], Hofmann and Rüsch in 2017 [48], Xu et al. in 2018 [15], Oztemel and Gursev in 2018 [17], and Kosacka-Olejnik and Pitakaso in 2019 [13].
"Supply Chain 4.0" or "Smart Supply Chain" is a term that declares the transformation of a traditional SC when Industry 4.0 technologies are adopted [11,45,59,80]. There are numerous studies that analyze the impression of Industry 4.0 on SC management and logistics . The study presented in [56] by Sharma et al., shows that Industry 4.0 technologies can support the implementation of sustainable SC management initiatives in multi-tier manufacturing SCs. The results of the study conducted by Tjahjono et al. [59] showed that the areas which are most affected by the introduction of Industry 4.0 technologies in SC are order fulfillment and transport logistics. Garay-Rondero et al. [46] proposed a conceptual framework of digital SC, which incorporates the role of Industry 4.0. Moreover, there are many studies that focus on the impact of a single Industry 4.0 technology on SC management and logistics [61][62][63][64][65][66][67][68][69][70][71][72][73][74][75][76][77][78][79].
A "Resilient Supply Chain 4.0" results when Industry 4.0 technologies are applied to a SC aiming to enhance its resilience. The literature related to "Resilient Supply Chain" is mainly focused on the determination of the constituent elements of SCRes, their interrelation, and their grouping. The literature related to "Resilient Supply Chain 4.0" investigates the impact of Industry 4.0 technologies on SCRes or on the constituent elements of SCRes.
A recent study conducted in 2022 by Tortorella et al. [80] identified the contribution of Industry 4.0 technologies into SCs with the objective of improving SCRes. Several studies give evidence that Industry 4.0 technologies positively impacts SCRes [84,95,119], while Sharma et al. [56] found that resilient practices are positively associated with Industry 4.0 technologies. Dilyard et al. [82] examined the use of Industry 4.0 technologies to enhance SCRes in the case of the COVID-19 pandemic.
Ivanov, Dolgui, and Sokolov [91] have studied how Industry 4.0 affect the SC disruption risk management and the ripple effect to design a resilient supply chain. The ripple effect occurs when a disruption cannot be prevented at a local level but is propagated downstream impacting SC performance. Valilai and Sodachi [60] consider that resiliency is embedded in sustainability assessment. They propose a sustainability assessment model that connects the sustainability assessment frameworks with Industry 4.0 related analytical platforms, with the help of the resiliency concept. Table 4 includes the selected Industry 4.0 technologies, which are found to influence SCRes according to the literature review. In the following, we will discuss each of these technologies and report how they influence each of the selected SCRes elements that are included in Table 3. The results of this investigation are then included in Table 5.

Internet of Things (IoT)
The term IoT was introduced by Kevin Ashton in a presentation that he gave at Procter & Gamble in 1999. Ten years later, Ashton [23] comments that the economy, society, and survival of humanity are mainly based on things, not on ideas or information. IoT is considered one of the core components of Industry 4.0. Connected via the Internet, smart things communicate with each other at any time and at any place, and exchange data using radio frequency identification (RFID). In a SC management context, things can be palettes, containers, or packages that can be automatically directed through logistics systems. The transport of goods can be controlled using RFID tags with appropriate sensors, such as temperature sensors for frozen goods and vibration sensors for fragile goods [50]. RFID can improve freight security and visibility, since cargo transportation is tracked and monitored [3]. When IoT is applied to industry, it is called Industrial Internet of Things (IIoT) [5,19].
Examining the impact of IoT on the elements of SCRes, studies and reports have provided evidence for most of the elements. Firstly, IoT has an effect on flexibility, enabling quick SC reconfiguration, and rapid recovery from disruptions through the efficient utiliza-tion of information. [98]. IoT can improve the visibility of inventory positions and logistics flows [47,80,98]. Moreover, IoT improves agility by increasing the speed of information flow [47,98]. IoT enables information sharing across multiple tiers [47,51,80] and improves collaboration by increasing trust among partners due to the acceleration of information sharing [47,96,98]. Then, IoT influences the management of knowledge within organizations, since it enables the collection and exchange of internal and external data [24]. Finally, other SCRes elements that are reported to be enhanced by IoT are: situation awareness [96], SCRM culture [96], security [98], robustness [47], risk management [47], and velocity [96] (see Table 5).

Cyber-Physical Systems (CPSs)
CPSs are considered among the main drivers of Industry 4.0. CPSs are systems that integrate both cybernetic and physical components allowing manufacturing equipment to embed computing power [63]. The cybernetic network consists of linked intelligent controllers, and the physical network consists of interconnected components of the infrastructure [10]. CPSs achieve the integration using IoT, RFID and sensor networks [26]. CSPs transform the operation of many traditional industrial systems, improving the effectiveness and efficiency of the entire industry [16]. The impact of CPSs on the SC is related to the high level of connectivity, intelligence, and automation [63]. A specific category of CPSs applied in the domain of manufacturing/production are Cyber-Physical Production Systems (CPPSs) [20]. CPPSs are smart production systems with embedded sensors and actuators, being aware of their condition and taking decisions without human intervention [5].
AI can be applied to provide a CPS with intelligent behavior. In case of machine damage, the machine agent decides the recovery method, such as overcoming the damage, cooperating with other machines to assign the work to an appropriate machine, or rescheduling [26].

Augmented Reality (AR)
AR uses electronic devices to view a real-world physical environment combined with virtual elements in real time [28]. AR gadgets and equipment include head up displays, helmet mounted displays, holographic displays, smart glasses, and handheld devices (cell phones and tablets) [29]. AR can be applied to service, manufacturing, sales and marketing, design, operations, quality affirmation, and staff training and support [28,29]. It must be clarified that Virtual Reality differs from AR. In Virtual Reality, a computer-assisted simulation tool uses multi-sensory channels to permit human-computer interactions and create a self-contained world that is disconnected from reality [36,64].
AR can be used for creating virtual models of the product during the development phase, allowing the detection of flaws without the physical prototypes [18]. This improves flexibility and the risk of damage is minimized [64]. Another example is the application of AR in staff training and support as real-time information, which can be provided either at the working environment or remotely [18,64]. Therefore, collaboration among partners is facilitated [84]. Concerning knowledge management, del Amo et al. [30] has shown that AR can be used to support three main knowledge management activities: knowledge transfer, knowledge capture, and knowledge discovery.
Concluding the impact of AR on the elements of SCRes, the literature review provides evidence that AR contributes to the enhancement of flexibility [64], collaboration [84], knowledge management [30], and velocity [47] (see Table 5).

Cloud Computing (CC)
CC services permit on-demand network access to computing resources, which are remotely provided by an internet service provider [18]. The combination of IoT with CC permits the implementation of wireless equipment connection and information sharing. In other words, different machines, devices, or items can be connected and exchange information in real time, resulting in enhanced decision making [18]. CC is also utilized for big data storage and process enabling BDA.
Arsovski et al. [121] proposed a model for assessment of organizational resilience potential in small and medium enterprises of the automotive SC, when cloud technology is used for logistics management, database management, and forecasting and planning demands.

Internet of Services (IoS)
IoS enables companies to provide their services through the Web technologies to make services easily accessible [10]. Therefore, IoS permits the offering of internal and cross organizational services to all participants of the value chain, with the support of CC and BDA [12,31,47]. IoS consists of four different parts: services, business models, participants, and infrastructure [46,47]. Users can access services via various channels. According to Reis and Gonçalves [31], IoS, IoT, and CPS are the pillars of Industry 4.0 that can support the creation of a smart factory.

Big Bata Analytics (BDA)
BDA provides analysis and categorization of massive volumes of data into useful information and knowledge, which can support the decision-making processes in organizations [104]. Since BDA enables access to real-time critical information, organizations can exploit this information to act proactively and reduce disruption risks [105,106]. Moreover, BDA can trace the origin of disruptions, monitor disruption propagation, improve risk management, and be used to best achieve SCRes [84]. Yamin [155] gave evidence that BDA has a moderating impact on SCRes and sustainable SC performance. Although BDA improves SCRes, recent studies reveal that BDA is relatively low when implemented by manufacturing companies [18,70].

Artificial Intelligence (AI)
Baryannis et al. [70] state that the key approaches in AI include: (a) mathematical programming approaches (stochastic programming, fuzzy programming, robust optimization, and hybrid approaches of various forms of mathematical programming); (b) network-based models (Petri Nets and Bayesian networks); (c) agent-based modeling and multi-agent systems; (d) automated reasoning; and (e) Machine Learning and BDA. It is worth mentioning that some researchers consider BDA to be a part of AI [70], while some consider AI and BDA as two separate branches of computer science that complement each other [33,91,96]. Other researchers have focused on the impact of Machine Learning [80,84], agent-based simulation [154], and Dynamic Bayesian Networks [146] on SCRes.
The findings of various studies indicate that AI techniques have a positive impact on SCRes [70,83,107]. According to a survey conducted in 279 firms by Belhadi et al. in 2021 [107], the information processing capabilities of AI can be exploited to develop SCRes for enhancing SC performance. In particular, the results indicate the mediating role of collaboration and adaptive capacity between AI and SCRes. Ali et al. [120] argue that increasing flexibility or building redundancy can improve the ability of companies to adapt to changes in case of disruption.

Digital Twins (DT)
Ivanov and Dolgui [109] identify a DT supply chain as "a computerized model that represents network states for any given moment in real time". The physical SC is mirrored in a virtual SC using advanced technology. DT constitutes an innovative solution for building a smart SC [72]. The main technologies that are utilized to create a DT from a physical SC are simulation and optimization combined with data analytics [109]. In addition, CPS, RFID, IoT, and BDA are among the key technologies of DT supply chain [109,147]. The DT model proposed by Ivanov and Dolgui [109] can be used in risk management. Kalaboukas et al. [108] propose the concept of Cognitive DT for SC management, which can predict trends and can also be flexible in dynamic environments.

Blockchain (BC)
BC technology can be considered as a distributed database where ledger data is stored and shared among participants from different organizations. Participants agree in advance the rules of data sharing [73]. All data shared are performed in an encrypted version. "Block" is used to name each new set of transactions added to the platform. Bitcoin and Ethereum are famous examples of public BC platforms with open access [114]. There are also private BCs and consortium BCs that combine public and private BCs, permitting access to pre-defined participants [73,114].
BC can directly connect all partners and systems in the SC and enhance SCRes by exploiting accurate and real-time data to make decisions [85]. Weber et al. [76] proposed the adoption of BC technology for increasing trust in collaborative business processes. Risk management in the SC is influenced by BC since secure information sharing is facilitated almost in real-time [110]. Manupati et al. [158] propose the application of BC technology to predict the time of the occurrence of a disruption in the SC network. Smart contracts can be used to monitor the response of merchandise assets to the uncertainty. A threshold limit is used to detect if the disruption is complete or partial and proceed to proper actions. Sreenu et al. [1] propose the integration of IoT sensors with BC to monitor temperature in vaccine SC. BC records any temperature violation as a transaction that cannot be changed. In such a case, the purchase order is rejected by the buyer.
Since BC security problems have become known recently, increasing security is a need [139]. Shrivastava et al. [27] have discussed the connection between AI and BC to provide defense against cyber threats. A recent literature review conducted by Bayramova et al. in 2021 [114] found that BC has the potential to enhance SCRes to cybercrime. The findings of a study reported by Lohmer et al. [110] revealed that BC improves SCRes, reducing the recovery time, the cost of disruption, and the number of affected partners. The researchers [110] found that if BC is poorly implemented, then short disruptions have a strong negative impact on SCRes. Similarly, Hervani et al. [159] argue that the firms' resilience performance relies on the implementation of BC internally and across their SCs.

Industrial Robotics (IR)
IR can take advantage of AI, BDA, and CC to become more intelligent, resulting in a complete transformation of the production and manufacturing phase [38]. In the Industry 4.0 context, IR have integrated sensors which collect data in real-time and upload data to the cloud computers. Data is stored, traced, and analyzed [93], and then utilized to enable sensing, comprehension, acting, and learning [38]. A special category of robots are collaborative robots or cobots, which help workers perform their tasks, using different modes of collaboration. Cobots can make workers more effective and efficient [36].

Additive Manufacturing (AM)
AM (or 3D printing) is used for the construction of 3D products by adding layers of materials on top of each other. AM can be used to facilitate rapid and on-demand production [93]. AM is utilized by different sectors, such as aerospace, automotive, construction, medical, and consumer goods [36]. Companies can exploit AM to accelerate the production of several prototypes and complex parts, which can then be used in the production process.
AM provides flexibility in design and operation, since different components and products can be produced in the same production line, achieving the customization of products when this is demanded [18,91,116,151]. Moreover, AM can reduce the number of SC layers and suppliers [91].

Industry 5.0 Technologies
It is worth mentioning that even though organizations have not fully integrated Industry 4.0, the recent European Commission announcement of the fifth Industrial Revolution, or Industry 5.0 [162], has created an increased worldwide interest. Many key enabling technologies support the transition from Industry 4.0 to Industry 5.0, such as BDA, DT, BC, AI, and Cobots (collaborative robots) [163]. In Industry 4.0, these technologies are mainly utilized to fulfil the objective of the technological evolution of manufacturing and production systems. Industry 5.0 goes further, focusing on human-centricity, sustainability, and resilience [162][163][164]. In other words, Industry 5.0 respects worker needs, sustainability, and environmental harmony, plus the need to enhance resilience across value chains and supply chains [163].
New advanced technologies can support Industry 5.0, such as the Internet of Everything (IoE), Edge Computing (EC), eXtended Reality (XR), and 5G and Beyond [163]. IoE extends IoT, connecting not only physical objects but also persons, organizations, and systems. EC allows for data processing at or near the user where data is being generated. Examples of edge devices are self-driving vehicles, autonomous robots, and sensors. XR is a term which includes AR, Virtual Reality (VR), and Mixed Reality (MR). 5G and 6G networks are expected to support future high-quality services of Industry 5.0.

Determining KPIs for SCRes
Key Performance Indicators (KPIs) are criteria that enable the evaluation of the performance of processes of operations in any organization. Decision #makers continuously monitor KPIs, aiming to improve an organization's performance. KPIs can also be used for the identification of the origin of a problem.
In Section 2, we discussed the literature review approaches that have appeared by searching the Scopus database for publications investigating the impact of Industry 4.0 technologies on KPIs used for creating a resilient SC. Table 1 depicts the most relevant publications [92,117,146,149,150] of our investigation in the Scopus database. In the following part, we discuss some of the non-financial KPIs that are referred to these publications.
Dev et al. [92] show that the manufacturing on an AM machine improves SCRes in case of a disruption in the diffusion of green products, increasing the service level (or fill rate), increasing the recovery speed, and decreasing the stock level.
Patidar et al. [117] evaluated KPIs for SCRes analyzing expert views with fuzzy analytic hierarchy process. The results included four KPI categories of a resilient SC: time oriented, operational, structural, and organizational. Time-oriented KPIs include lead time, time to recovery, and order cycle time. These KPIs are widely used to measure the SCRes. Operational KPIs include capacity utilization, risk assessment frequency, supplier delivery efficiency, supplier rejection rate, SC cycle time, and demand and supply variations. Structural KPIs include the number of SC nodes and the proximity to suppliers and customers. Finally, organizational KPIs include service rate, equipment effectiveness, inventory velocity, stock level, and forecasting accuracy and errors.
Ansari and Kohl [146] propose an AI-enhanced approach that can improve KPIs for SCRes, such as Remaining Useful Lifetime, Mean-Time Between Failure, and Uptime, which have an impact on Overall Equipment Efficiency of production systems.
Catellani and Bottani [149] conducted a survey for evaluating the usage of 112 metrics to measure SC performance. Among the KPIs for SCRes that were asked to be evaluated are safety stock, inventory gap, recovery time, on-time delivery, and customer service level. The KPI that was used by any company that participated in the survey was the on-time delivery. The authors [149] evaluated the impact of Industry 4.0 technologies on SC performance, but they did not determine the relation of Industry 4.0 technologies with each of the selected KPIs.
Dumitrascu et al. [150] developed a SC management performance evaluation model that links specific problems with the most relevant KPIs. The authors used data mining to discover the relationships between SC management subsystems and specific KPIs. The results show the top five KPIs for every one of the nine subsystems of SC management. For example, safety stock is related to warehouse management, production lead time is related to production management, SC cycle time is related to demand management, while perfect order delivery rate is related to distribution management.
Moreover, we have investigated approaches that identify KPIs for enhancing SCRes, without the restriction of the Industry 4.0 perspective. The results mainly contribute to the enhancement of the comprehension of the KPIs that we have already mentioned.
Karl et al. [118] identify thirteen elements of SCRes and categorize them into three phases: before disruption, during disruption, and after disruption. They then analyze the relationship of SCRes elements with 10 non-financial KPIs of organizations. For example, the supplier rejection rate is a KPI that requires the collaboration of SC partners to improve its operational results. Karl et al. [118] propose two indicators related with the customer: consumer satisfaction and damage return rate. While customer satisfaction is revealed to be related to many SCRes elements, damage return rate is revealed to be connected only with collaboration in the during-disruption phase [118]. The authors [118] found that two SCRes elements, i.e., financial strength and visibility, have no link to any of the KPIs, and we have adopted these results. Moreover, an interesting result [118] is that the knowledge management element has the highest co-occurrence in literature with the identified KPIs.
Rajesh [160] considers five elements of SCRes: flexibility, responsiveness, quality, productivity, and accessibility. For each of these elements, Rajesh [160] proposes several measures. For example, an important responsiveness indicator is on-time delivery ratio. Productivity indicators, which are also related with customer satisfaction, include order compliance, fill rate, storage space utilization, and units moved per person-hour.
Singh et al. [161], propose a SCRes framework where they identify KPIs for each of the SCRes constituent elements. For example, collaboration, information sharing, and awareness are SCRes elements that can help achieve an accurate forecast demand. Velocity is another SCRes element that can improve the speed of recovery.
Johnson et al. [165] differentiate the robust SC from the resilient SC. A robust SC can return to near normal operation after a disruption of low to medium severity, while a resilient SC can return to normal operation even after a disruption with high severity. The authors [165] test the effectiveness of four risk mitigation strategies for SCRes in case of disruptions with various severity levels, using KPIs such as order fulfillment rates and profits. They found that flexibility and strategic capacity, i.e., the spare capacity at each facility, are the most successful strategies for moderate and severe disruptions. The proposed risk mitigation strategies for a robust SC are node redundancy and supply redundancy. Increasing the number of SC nodes results in decentralization being achieved, which means that if a node fails, total failure is avoided and robustness is maintained.
Rahman et al. [166] evaluate the most critical KPIs for SCRes and sustainability in the healthcare supply chain in the context of the COVID-19 pandemic. The results show that in the case of unusual disruptions, the KPIs, which are more crucial, are first the SC planning and forecasting performance, next the delivery and distribution performance, and finally the manufacturing and production performance as well as the procurement and sourcing performance.
Carvalho et al. [167] propose a framework for modeling the resilience of on-time delivery to capacity and material shortages. They argue that the reduction of recovery time is achieved by improving flexibility, collaboration, and responsiveness.
Lamballais et al. [168] model and analyze the performance of robotic mobile fulfilment systems. The proposed models estimate robot utilization, workstation utilization, and average order cycle time. The findings of this study [168] also show that the location of the workstations around the storage areas affects the maximum order throughput.
Kinra et al. [169] use KPIs of risk inclusion in the SC, focusing on supplier vulnerability. They propose a supplier risk exposure model, which can be used to quantify the ripple effect of a disruption occurring on the supplier side. KPIs, such as supplier reliability, time to survive, and fill rate are used to calculate the supplier risk exposure and the ripple effect exposure of the supply chain.
Li & Zobel [170] identify KPIs for SCRes in case of risk propagation. The authors [170] propose three dimensions of resilience: robustness, recovery time, and average functionality. Two types of robustness are considered, at initial impact and at full impact. Then they propose specific structural KPIs, such as the number of healthy nodes, which is the number of non-disrupted nodes, and the size of the Largest Connected Component (LCC), which is the size of the largest connected subnetwork after the occurrence of a disruption. The authors [170] also argue that increasing the number of SC nodes can lead to higher robustness, while the recovery time increases, too.
Stevenson and Spring [171] argue that information sharing can help, among others, avoiding lost sales, avoiding over-production, and reducing inventories. Therefore, information sharing can help reducing supply and demand variations.
Considering the results of our investigation as presented above, we have formulated a list of non-financial KPIs used for building SCRes. In Table 6, the selected KPIs are listed. For each KPI, a description is given, and, in some cases, related KPIs found in the literature review are referenced. Table 6. KPIs for SCRes.

KPIs
Descriptions and References

Lead time
The time elapsed in between the placement of an order from a customer until the delivery to the customer [47,117,161] (Production lead time [150], Order lead time [118], Delivery lead time [92,118])

Time to recovery
The time needed for a facility to regain its usual level of functionality after a disruption [81,117,149,167,169,170] (Speed of recovery [92,120,161]) Order cycle time The time between two placed orders [117,168] Capacity utilization The ratio of the actual output to the maximum output [117,118] Risk assessment frequency Indicates how frequently risks are being identified [117] Supplier delivery efficiency Indicates how efficiently a supplier delivers the supplies [117,118] (Supplier reliability: The ratio of the quantity received over the quantity ordered [169])

Supplier rejection rate
The quantity of supplies rejected by the suppliers over the total quantity received [117,118] SC cycle time The sum of the longest lead time for each stage Indicates the efficiency of the supply chain [117,150] Demand and supply variations The variability between demand and supply [117,171] No of nodes in SC Number of product stops during the production process [117,165,170] Proximity to suppliers and customers The proximity between company and suppliers and the proximity between company and customers [117] Service rate The level of service to the customers [47,117,118,149,160,165] (Order fill rate: The ratio of the quantity shipped to the customer over the total quantity ordered [160,165]) (Customer satisfaction [47,118])

On-time delivery
The ratio of the number of orders delivered on date over the total number of orders delivered [118,148,149,160,167,169] Equipment effectiveness /efficiency The effectiveness/efficiency of the equipment in a manufacturing plant [117] (The ratio of fully productive time to planned production time [146])

Inventory velocity
The time interval between supplying raw materials and selling finished products [117] Stock level The optimum stock level [47,92,117,118,149,150] Forecasting accuracy The deviation of the actual demand from the forecasting demand [47,117,161,166]

Determining the Impact of Industry 4.0 Technologies on KPIs for a Resilient Supply Chain 4.0
In Section 4, we investigated the influence of Industry 4.0 technologies on selected SCRes elements. The findings are included in Table 5.
In Section 5.1, we investigated the KPIs for SCRes and the results are included in Table 6. In this Section, we report how each of the Industry 4.0 technologies is related with each of the KPIs, considering the literature review. The results of our investigation are then synthesized and aggregated in Table 7. For each KPI in Table 7, the findings include: (a) the Industry 4.0 technologies that have a direct impact on the KPI, and/or (b) the Industry 4.0 technologies that have an indirect impact on the KPI through the improvement of the SCRes elements which are related with the KPI.
Since Table 7 contains many repetitions for the sake of references, we have gathered the findings in Table 8. There, the matches show which Industry 4.0 technology improves which KPI of SCRes.
In the following, a discussion of the literature review findings is made by technology. Tables 7 and 8 then aggregate these findings.

Internet of Things (IoT)
IoT can help reducing the time spent by workers in filling and shipping orders and can facilitate tracking products and services [98]. Moreover, IoT can provide accurate information when a disruption event is observed and enable immediate handling [119]. As a result, IoT provides the benefit of shorter lead times [47], shorter time to recovery [119], shorter order cycle time, and shorter SC cycle time by increasing time efficiency in production [47], better service rate [47], better equipment effectiveness by reducing downtime [47], increased inventory velocity by inventory picking process improvement [47], and better forecasting accuracy [47] (see Table 7).
Furthermore, IoT can contribute to the enhancement of SCRes elements such as flexibility, agility, collaboration, information sharing, situation awareness, SCRM culture, risk management, knowledge management, and velocity, which are improved by Industry 4.0 technologies (see Tables 5 and 7). Table 7 shows which of these SCRes elements are related with each of the following KPIs: lead time [117,118,161], time to recovery [161,167], capacity utilization [118], risk assessment frequency, supplier delivery efficiency [118], supplier rejection rate [118], demand and supply variations [171], no of nodes in SC [117,170], service rate [118], on-time delivery [118,167], stock level [117,118], and forecasting accuracy [117,161] (see Tables 7 and 8).

Cyber-Physical Systems (CPSs)
CPS can reduce waste and work in lean manufacturing, improve maintenance, and increase traceability and transparency in distribution [47]. Therefore, CPS can reduce lead time [47], reduce order cycle time and SC cycle time by increasing time efficiency in production [47], and improve equipment effectiveness by reducing downtime [47] (see Table 7).
Additionally, AR can enhance SCRes elements such as flexibility, collaboration, knowledge management, and velocity, which are improved by Industry 4.0 technologies (see Tables 5 and 7). Table 7 shows which of these SCRes elements are related to each of the following KPIs: lead time [118,161], time to recovery [161,167], capacity utilization [118], risk assessment frequency, supplier delivery efficiency [118], supplier rejection rate [118], service rate [118], on-time delivery [118,167], and stock level [118] (see Tables 7 and 8).

Cloud Computing (CC)
CC can reduce SC cycle time by increasing efficiency [47], improve service rate by improving customer satisfaction [47], reduce the stock level [47], improve equipment effectiveness by reducing downtime [47], increase inventory velocity by improving the inventory picking process [47], and improve forecasting accuracy [47] (see Table 7).
Furthermore, CC can enable the enhancement of SCRes elements such as flexibility, collaboration, agility, information sharing, risk management, and velocity, which are improved by Industry 4.0 technologies (see Tables 5 and 7). Table 7 shows which of these SCRes elements are linked with each of the following KPIs: lead time [118,161], time to recovery [161,167], capacity utilization [118], supplier delivery efficiency [118], supplier rejection rate [118], demand and supply variations [171], service rate [118], on-time delivery [118,167], stock level [118], and forecasting accuracy [117,161] (see Tables 7 and 8).

Internet of Services (IoS)
IoS can reduce SC cycle time by increasing efficiency [47], improve equipment effectiveness by reducing downtime [47], and improve forecasting accuracy [47] (see Table 7).
Moreover, IoS can enhance SCRes elements such as flexibility, robustness, and risk management, which are improved by Industry 4.0 technologies (see Tables 5 and 7). Table 7 shows which of these SCRes elements have a relation with each of the following KPIs: lead time [118], time to recovery [167], capacity utilization [118], supplier delivery efficiency [118], no of nodes in SC [117,170], and on-time delivery [118,167] (see Tables 7 and 8).

Big Bata Analytics (BDA)
A study conducted in 2022 by Iftikhar et al. [106] investigated the mediating effect of BDA between SC complexities and SCRes. Two types of SC complexities are considered: structural complexities (e.g., no of SC nodes, products, parts, suppliers, customers) and dynamic complexities (e.g., lead times from suppliers, supplier delivery efficiency). The results showed that BDA has a mediating effect for structural and dynamic complexities on SCRes. Moreover, dynamic SC complexity only influences SCRes via BDA. According to Zouari [84], BDA can facilitate the tracing of the origin of disruptions and monitor disruption propagation. For this reason, BDA contributes to the reduction of recovery time, since accurate information concerning the occurrence of a disruption is available in time [106,119]. BDA can also reduce SC cycle time by increasing efficiency [47], improve equipment effectiveness by reducing downtime [47], and increase inventory velocity by improving the inventory picking process [47] (see Table 7).
In addition, BDA can contribute to the improvement of SCRes elements such as SC configuration, redundancy, flexibility, collaboration, agility, information sharing, situation awareness, robustness, risk management, knowledge management, and velocity, which are improved by Industry 4.0 technologies (see Tables 5 and 7). Table 7 includes the SCRes elements that are related to each of the following KPIs: lead time [117,118,161], time to recovery [161,167], capacity utilization [117,118], risk assessment frequency, supplier delivery efficiency [118], supplier rejection rate [118], demand and supply variations [171], no of nodes in SC [117,170], service rate [118], on-time delivery [118,167], stock level [118], and forecasting accuracy [117,161] (see Tables 7 and 8).

Artificial Intelligence (AI)
AI methods can be used for the proactive response to changes [80,84,146]. Examples of the utilization of AI methods are the management of urban freight transportation using agent-based simulation [154], and route optimization [117]. AI thereby contributes to the increase in proximity to suppliers and customers through real-time route optimization [117,154] and can also provide accurate forecasting [117]. Machine Learning (a key approach of AI [70]) can improve time to recovery by providing accurate information about the occurrence of a disruption in time [119], increase capacity utilization [47], improve inventory velocity [47], improve service rate [47], improve equipment effectiveness by reducing downtime [47,146], and improve forecasting accuracy [47] (see Table 7).

Digital Twins (DT)
DT can help to minimize lead times [108], reduce SC cycle time by increasing efficiency [47], increase service rate by improving interaction with the customer [47,147], reduce equipment effectiveness by reducing downtimes [47], improve equipment effectiveness by reducing downtime [47], increase inventory velocity by improving inventory picking process [47], and improve forecasting accuracy [47,108] (see Table 7).

Blockchain (BC)
BC can contribute to the increase in supplier delivery efficiency [85], the decrease in demand and supply variations by reducing lost demand [85], the reduction of stock level [85], and the increase of forecasting accuracy [85] (see Table 7). BC also contributes to the reduction of lead time [117,119], the significant shortening of recovery time [110,117,119], and the reduction of order cycle time and SC cycle time [117] (see Table 7).
Additionally, BC can help enhance SCRes elements such as SC configuration, flexibility, agility, collaboration, information sharing, SCRM culture, knowledge management and velocity, which are improved by Industry 4.0 technologies (see Tables 5 and 7). Table 7 lists the SCRes elements that are related to each of the following KPIs: lead time [118,161], time to recovery [161,167], capacity utilization [118], risk assessment frequency, supplier delivery efficiency [118], supplier rejection rate [118], demand and supply variations [171], service rate [118], on-time delivery [118,167], and stock level [118] (see Tables 7 and 8).

Industrial Robotics (IR)
IR can contribute to the reduction of order cycle time (the performance of a robotic fulfillment system is reported in [168]), the reduction SC cycle time [36,47], the increase in equipment effectiveness by downtime reduction [47], the increase in inventory velocity by inventory picking process improvement [47], and the minimization of stock level [47] (see Table 7).

Additive Manufacturing (AM)
AM can help to reduce lead time [47,91,116], decrease time to recovery [92], increase capacity utilization [47], reduce order cycle time [116], reduce SC cycle time by increasing resource efficiency [47], increase equipment effectiveness by reducing downtime [47], increase service rate [92], and reduce stock level since fewer items need to be held [91,92,116,151] (see Table 7).
In addition, AM can enhance SCRes elements such as SC configuration, redundancy, flexibility, agility, robustness, and velocity, which are improved by Industry 4.0 technologies (see Tables 5 and 7). Table 7 shows which of these SCRes elements are related to each of the following KPIs: lead time [117,118,161], time to recovery [92,167], capacity utilization [117,118], no of nodes in SC [117,170], service rate [118], on-time delivery [118,167], and stock level [118] (see Tables 7 and 8).

Conclusions
This study aims to extend current research results on how the KPIs for making a resilient SC are influenced by the most important Industry 4.0 technologies. In order to fulfill this objective, we have examined two research questions. The first research question (RQ1) sought to find which of the Industry 4.0 technologies can improve SCRes. Since the impact of Industry 4.0 technologies on SCRes is not straightforward, we explored how each of the Industry 4.0 technologies influence the most important SCRes constituent elements, and we discussed the findings. The results are aggregated in Table 5. Therefore, we answered RQ1.
The second research question (RQ2) sought to investigate which of the Industry 4.0 technologies influence the KPIs for a resilient SC. We followed a systematic literature review method to answer RQ2. The search results in Scopus database were very limited, and so we extended our literature review. We studied the literature findings, and we decided to mainly adopt Patidar's et al. [117] list of non-financial KPIs for a resilient SC, enhanced with more literature references. These KPIs are depicted in Table 6. For each KPI appeared in Table 6, a two-fold research process was followed. Firstly, we investigated the link between the SCRes constituent elements and the adopted KPIs. In other words, for each KPI, the related SCRes constituent elements were determined according to the literature review. We synthesized these findings with the findings of RQ1 (depicted in Table 5), and we concluded which Industry 4.0 technologies influence each KPI. Secondly, based on the literature review, we determined which Industry 4.0 technologies have a direct impact on each KPI. For each Industry 4.0 technology, we presented the related KPIs and we depicted the detailed results in Table 7. For reasons of simplicity and comprehension, the findings shown in Table 7 were then aggregated in Table 8.
Our study is differentiated from other studies. Among the results of the systematic literature review, only Patidar et al. [117] is closer to our study. The authors [117] evaluated KPIs for SCRes and investigated the influence of Industry 4.0 technologies on these KPIs based on the literature review. However, we have extended this research investigating the impact of Industry 4.0 technologies on the constituent elements of SCRes (such as Govindan et al. [47]) and the relation of the KPIs for SCRes with the constituent elements of SCRes (such as Karl et al. [118]). Other studies are concentrated on the impact of a particular Industry 4.0 technology on KPIs for SCRes (e.g., AM [92]), while others evaluate the impact of Industry 4.0 technologies on SC performance without determining the relation of Industry 4.0 technologies with each of the KPIs (e.g., [149]).
The findings of this work could be exploited by academics and practitioners who are interested in the concept of a Resilient Supply Chain 4.0. The Industry 4.0 technologies that are revealed to improve the KPIs for creating SCRes could be considered for adoption by companies of various industries who intend to improve the resilience of their SC.

Limitations and Future Research
This study has some limitations. At first, the study relied on high quality articles published in journals that were mainly searched in Google Scholar and Scopus databases. As a result, there is a possibility that some relevant articles are not included. A second limitation is that we have based our research on results presented by other researchers. For the future, we plan to conduct an empirical investigation in order to identify the KPIs that help measure resilience in Supply Chain 4.0 and then evaluate these KPIs using a decision support method. The impact of Industry 4.0 technologies on the evaluated KPIs could be investigated using case studies and interviews with experts.
Figures and tables were used for the visualization of our work on the literature review and the results of this research. Advanced tools such as VOS Viewer [172] can be applied to construct and analyze bibliometric networks and provide a clear depiction of item connections.
An interesting future work would be the investigation of the influence of the Industry 4.0 maturity models on the KPIs for SCRes. Maturity models of Industry 4.0 are used for auditing any company, measuring its progress, and comprehending its strengths and weaknesses [173]. Taking into consideration that the introduction of Industry 4.0 in enterprises, especially in SMEs, should be implemented systematically, maturity models can be useful tools [174,175]. Different maturity models of Industry 4.0 are found in the literature [176].
For the future, the business society should take into account that the 4th and the 5th Industrial Revolutions co-exist, where Industry 4.0 is considered technology-driven and Industry 5.0 is considered value-driven [164]. Resilience is one of the core values of Industry 5.0, which is interconnected with human-centricity and sustainability [163,164]. The new emerging digital technologies such as individualized human-machine interaction, bio-inspired technologies, smart materials, and innovations in AI [164] will force industries to continually transform products and processes [17].