Building antifragile manufacturing systems through strategic technology integration Open Access

Interesting because – Manufacturing firms face persistent turbulence that challenges traditional approaches to stability and recovery. This study is noteworthy because it reframes disruption as a driver of progress. It introduces antifragility as a strategic capability that enables organizations to improve through variability rather than resist it. By linking digital transformation to capability development, the research explains how sensing, seizing, and transforming can operate as an integrated system that strengthens with each cycle of change.

Theoretical value – The study extends dynamic capability theory by formalizing antifragility as a higher-order construct that unites technological adaptation with continuous learning. It distinguishes operational responsiveness from operational dynamism and explains how their interaction drives cumulative capability growth. The results demonstrate how technologies such as IoT, AI, and analytics foster self-reinforcing improvement, transforming variability into structured renewal and long-term adaptability.

Practical value – For managers, the framework provides actionable guidance for aligning digitalization with antifragile growth. It helps prioritize which capabilities to develop, design governance that promotes cross-functional coordination, and select technologies that enhance learning and flexibility. The approach encourages firms to treat disruptions as learning cycles, use technology adoption strategically, and build systems that improve performance, responsiveness, and competitiveness over time.

In today’s dynamic and uncertain environment, manufacturing systems face recurrent disruptions ranging from supply chain volatility to technological shifts and macroeconomic shocks (Ivanov, 2023a). Established disorder response approaches increasingly fall short in enabling firms not only to withstand shocks but to improve because of them (Nikookar et al., 2024). This has renewed interest in antifragility, a capacity to benefit from variability that remains underdeveloped in manufacturing research and practice (Becker et al., 2024). Antifragility is multifaceted and strategic, requiring cohesive capabilities that operate across the organization rather than isolated tools or ad hoc fixes (Hillson, 2023). Digitalization intensifies the need for such cohesion. Firms of all sizes invest in artificial intelligence (AI), the Internet of Things (IoT), data analytics, and automation to enhance efficiency and adaptability (Fosso Wamba et al., 2024), yet many initiatives are not aligned with antifragility aims. Misalignment risks missed opportunities and can even increase fragility (Roy et al., 2024).

Recent studies have advanced understanding of how manufacturers can survive and even thrive under turbulence (Javed et al., 2024a, b; Rashid et al., 2024). For instance, evidence shows that supply chain integration and knowledge capabilities, when aligned with technological dynamism, foster ambidextrous innovation in manufacturing (Javed et al., 2024a). Research on pandemic-related disruptions indicates that combining advanced technologies with supply chain collaboration enhances sustainable supply chain performance (Javed et al., 2024b). Complementary work suggests that innovation and knowledge-based capabilities, supported by technological capabilities, can strengthen greener and higher-performing manufacturing (Javed et al., 2025). However, the literature does not yet conceptualize antifragility as a distinct capability system that improves through volatility, nor does it provide an implementable mapping that links digitalization to this system across sensing, seizing, and transforming while clarifying organizational boundaries in relation to resilience and viability. Addressing this gap is the motivation for the present study.

Theoretical gaps therefore persist in specifying the capabilities that underpin antifragility in manufacturing (Becker et al., 2024; Priyadarshini et al., 2022). Foundations in resilience and adaptability are relevant yet insufficient because they prioritize resistance and recovery. Antifragility treats volatility as a source of systematic improvement rather than only a risk to be managed (Ghobakhloo et al., 2025; Munoz and Zhou, 2023). The literature still lacks a cohesive account that defines antifragility as a distinct capability system, distinguishes it from resilience and viability, and maps how digitalization supports that system in volatile, complex manufacturing settings.

Practice reflects similar challenges. Many firms remain unsure how to operationalize antifragility strategically when integrating it with digital transformation (Giordino et al., 2024). Common problems include misaligned resources, weak cross-functional coordination, and difficulty embedding new tools into day-to-day work, which can leave investments costly yet ineffective (Ghobakhloo et al., 2023; Sony et al., 2022). Misalignment between digital initiatives and antifragile objectives can introduce operational vulnerabilities and inefficiencies (Ghobakhloo and Iranmanesh, 2021). For manufacturers seeking durable advantage in shifting markets, strategy must connect digital capabilities to an explicit antifragility capability set.

This study addresses these gaps in three steps. First, it defines antifragility for manufacturing as a distinct higher-order dynamic capability and specifies a coherent capability set across sensing, seizing, and transforming, clarifying its difference from resilience and viability (Becker et al., 2024; Ghobakhloo et al., 2025). Second, it develops a technology–capability map that links contemporary digitalization domains to those capabilities and identifies organizational catalysts required for reliable implementation at scale (Munoz and Zhou, 2023). Third, it validates and refines the framework through expert input to ensure conceptual clarity and practical relevance. Together, these steps translate a broad capability conversation into an implementable approach for manufacturing antifragility. Accordingly, the study strives to answers the following research questions:

This paper contributes to the literature by conceptualizing antifragility in manufacturing as a distinct higher-order dynamic capability and by specifying its underlying capability system across sensing, seizing, and transforming. It advances theory by clarifying how antifragility differs from resilience and viability and by integrating digital technologies as enablers of this capability system. From a practical perspective, the framework provides manufacturers with a structured approach to align digitalization with antifragility objectives, offering guidance on how to turn disruptions into opportunities for learning, adaptation, and long-term competitiveness.

Antifragility in manufacturing represents a distinct way of dealing with variability and disorder. Taleb (2012) defines antifragility as the capacity to improve because of stressors, volatility, and uncertainty. This stands apart from approaches that aim to resist change or recover after disturbances. Antifragile systems treat uncertainty as a driver of adaptation, innovation, and long-term advantage. The idea has gained prominence with advances in digital technologies and a growing recognition that dynamic environments can be harnessed for opportunity rather than only managed as threats (Becker et al., 2024; Nikookar et al., 2024).

To clarify this distinction, it is useful to compare antifragility with other disorder response systems. Table 1 outlines fragile, robust, resilient, viable, and antifragile systems according to their goals and expected responses to disruption. Fragile systems are highly vulnerable and deteriorate under stress, which is often the case in smaller firms with limited resources (Becker et al., 2024; Corvello et al., 2023). Robust systems prioritize stability by resisting change within design limits, which minimizes exposure but leaves little room for adaptation and may reduce competitiveness in shifting environments (Mondal et al., 2014; Srivastava et al., 2021). Resilient systems emphasize recovery, restoring operations after disruptions but generally aiming for a return to baseline rather than improvement (Kähkönen et al., 2023; Lerch et al., 2024; Ghobakhloo et al., 2025; Ivanov, 2023b). Viable systems extend resilience by balancing adaptability and sustainability to maintain long-term functioning, although their equilibrium orientation constrains the ability to turn volatility into growth (Ruel et al., 2024; Zekhnini et al., 2022; Ivanov and Dolgui, 2020).

Antifragility differs in both objective and logic. Instead of seeking stability, balance, or recovery, antifragile systems improve through disruption. In manufacturing this involves proactive mechanisms such as early sensing, experimentation, and continuous learning that convert shocks into innovation and renewal. Digital technologies, including real time analytics, predictive modeling, and digital experimentation, reinforce these mechanisms by enabling firms to transform variability into strategic advantage (Größler, 2020; Lotfi et al., 2023; Becker et al., 2024). While some practices such as redundancy or adaptability may appear across resilience, viability, and antifragility, the purpose they serve is not the same. In resilience they support recovery, in viability they secure long-term balance, and in antifragility they create the conditions for growth through volatility.

These categories should not be interpreted as stages on a continuum. They are distinct orientations, each suited to different contexts and objectives. The contrasts summarized in Table 1 highlight antifragility as a separate construct that moves beyond stability or recovery to focus on systematic improvement because of variability (Nikookar et al., 2021; Ruel et al., 2024).

The Delphi method was used because antifragility in manufacturing is still an emerging concept with limited empirical grounding. This method allows experts to iteratively refine definitions and relationships through structured, anonymous, and feedback-driven discussions (Beiderbeck et al., 2021). It was therefore well suited for identifying and validating the capabilities and technology linkages that underpin antifragile manufacturing.

Experts were recruited through professional networks, publication records, and industry reputation. Eligibility required at least ten years of experience in senior manufacturing or supply chain roles or multiple peer-reviewed publications on Industry 4.0, dynamic capabilities, or operational resilience. Fourteen experts were selected to ensure variation across industries, firm sizes, and roles, including plant managers, digital transformation leads, supply chain directors, and academic researchers. Anonymity was maintained to promote independent judgment and reduce social or hierarchical bias.

The Delphi process followed three iterative rounds adapted from Okoli and Pawlowski (2004). In Round 1, experts reviewed literature-based statements that defined candidate capabilities and technological enablers grouped under sensing, seizing, and transforming. Participants commented on clarity, scope, overlaps, and missing links. Peer comments were hidden to avoid anchoring. Round 2 provided a structured summary of the main points of agreement and contention, along with a visual draft of the capability–technology framework. Experts refined definitions, adjusted links, and proposed new relationships. Round 3 concentrated on unresolved items only, where participants voted to retain, revise, or remove each element while providing short justifications. Controlled feedback ensured that opinions evolved through reflection rather than persuasion.

Consensus thresholds were defined in advance. Agreement required at least 80% endorsement with no strong objections for definitions and 75% agreement for relationships. Items that remained stable between Rounds 2 and 3 were considered confirmed, while unresolved items were retained as contested. Qualitative inputs from all rounds were coded using structured content analysis. Two coders independently analyzed a subset of the data to develop a shared codebook that included affirmations, definitional refinements, merges or splits, new capability proposals, and links between capabilities and technologies. Discrepancies were reconciled, and the codebook was applied to the entire dataset to trace how expert input informed framework revisions.

Several measures supported validity and reliability. Anonymity, randomization of item order, hidden peer comments, and inclusion of minority opinions helped minimize bias. Seeding Round 1 with literature-derived definitions ensured comprehensive coverage of relevant topics while leaving room for refinement. Independent double coding enhanced interpretive accuracy. Robustness was further tested through sensitivity analyses, showing that excluding any single expert did not alter consensus outcomes, and subgroup comparisons indicated stable responses across professional backgrounds. By the conclusion of Round 2, the discussion had stabilized, and few new ideas or changes were proposed, indicating that the experts had reached a shared understanding of the key constructs and relationships. All participants provided informed consent and could withdraw at any time. Only aggregated findings were reported, and no confidential information was collected.

Antifragility, as conceptualized by Taleb (2012), refers to a system’s capacity not only to withstand stress and disruption but to benefit and grow stronger from them. In such systems, disruptions act as catalysts for improvement, fostering conditions where volatility and uncertainty fuel adaptation, learning, and long-term growth. In the manufacturing context, this study draws on the dynamic capability perspective (DCP) (Teece et al., 1997) as a robust theoretical foundation for translating antifragility into operational terms. DCP emphasizes an organization’s capacity to sense, seize, and transform in response to changing environments (Teece, 2023), thereby enabling the reconfiguration of resources and processes to adapt to or capitalize on variability (Wilhelm et al., 2015). These capabilities capture the essence of antifragility, outlining how manufacturing systems can evolve and improve in the face of fluctuating conditions.

The initial framework, shown in Figure 1, was constructed through an extensive review and synthesis of prior literature, identifying and defining the capabilities most critical for manufacturing antifragility and linking them to enabling technologies and governance mechanisms. These literature-derived constructs were then subjected to expert review and consensus building through the Delphi process described earlier. The Delphi engagement served to examine the clarity, relevance, and completeness of each capability definition and to refine the interrelationships between them based on practitioner and scholarly expertise.

DCP underscores three primary capabilities that allow firms to navigate complex and fluctuating environments strategically. Sensing refers to the ability to identify and interpret environmental changes, anticipating disruptions or opportunities (Kähkönen et al., 2023) through continuous monitoring and analysis (Ghosh et al., 2022). Seizing involves mobilizing resources and making strategic commitments to capitalize on opportunities or mitigate threats (Wohlleber et al., 2024), ensuring swift and effective reconfiguration of resources (Ghosh et al., 2022). Transforming encompasses the ongoing adaptation and renewal of a firm’s resource base in response to cumulative insights and environmental shifts (Kähkönen et al., 2023).

The sensing dimension includes adaptive learning and alertness, which equip systems to detect and interpret subtle environmental shifts (Größler, 2020; Nikookar et al., 2024). Adaptive learning builds on past experiences and integrates real-time data and insights from disruptions to improve processes over time (Gölgeci and Kuivalainen, 2020). Alertness enhances vigilance, enabling the detection of emerging patterns, weak signals, and nascent trends (Mandal, 2019). Proactive risk management anticipates potential disruptions using scenario analysis and contingency planning (Can Saglam et al., 2021; Hillson, 2023). These capabilities shift the organizational posture from reactive adjustments to proactive strategic positioning.

The seizing dimension enables manufacturing systems to respond dynamically to variability, with capabilities such as bricolage, which entails improvising with available resources to solve unexpected challenges (Essuman et al., 2023; Nikookar et al., 2024), and dynamic collaborations, which flexibly reconfigure relationships with suppliers, customers, and other stakeholders to secure complementary resources (Ghobakhloo et al., 2025; Ramezani and Camarinha-Matos, 2020). Operational responsiveness ensures rapid adaptation of production schedules, resource allocation, and supply chains in response to emerging conditions (Munir et al., 2022), supported by human and technological agility (Größler, 2020).

The transforming dimension embeds learning from disruptions into continuous improvement, with operational dynamism enabling processes and structures to evolve through the use of real-time data, analytics, and accumulated experience (Samadhiya et al., 2023). Operational responsiveness here supports the rapid reconfiguration of co-specialized resources to align with changing demands (Yu et al., 2019; Iyer et al., 2023). Self-improvement promotes decentralized decision-making and iterative learning, ensuring that each disruption strengthens organizational adaptability (Junaid et al., 2023; Hillson, 2023).

Figure 1 further illustrates how sensing, seizing, and transforming capabilities operate in a self-reinforcing cycle. Real-time data informs sensing capabilities, which detect and interpret changes; seizing capabilities mobilize resources to act on these insights; and transforming capabilities integrate lessons learned into structural and process improvements, which in turn enhance sensing over time. This continuous loop reflects the system-level nature of manufacturing antifragility, where the whole is greater than the sum of its parts.

Table 2 presents the final, expert-confirmed set of manufacturing antifragility capabilities, explaining how each capability’s function within the framework, emphasizing that the construct is both theoretically grounded in the DCP and empirically refined through structured expert consensus.

The enabling role of technology in manufacturing antifragility was established through a two-stage process: a synthesis of Industry 4.0 and resilience-oriented manufacturing research, followed by structured validation and refinement across the Delphi rounds. The literature review surfaced candidate technologies linked to sensing, seizing, and transforming; these links were then presented to the panel, who assessed relevance, clarified operational roles, and proposed refinements and additions. Table 3 reports the technology–capability links.

Across capabilities, IoT data, cloud integration, big-data platforms, and AI provide the shared information and analytics that support sensing, seizing, and transforming throughout the cycle. IoT-centric data capture provides continuous operational and supply signals, while cloud and big-data platforms enable integration, storage, and scalable analytics for monitoring, prediction, and coordination (Roy et al., 2024). AI converts these streams into forecasts, prescriptions, and decision aids that shorten detection and response times in sensing, seizing, and transforming (Chari et al., 2022; Ghobakhloo et al., 2025). Generative approaches are considered within AI because their effects arise through mechanism-specific contributions such as knowledge extraction, content generation, and scenario exploration rather than through a separate infrastructure layer (Ghobakhloo et al., 2024).

Within sensing, both the literature and the panel emphasized comprehensive data acquisition, real-time monitoring, and predictive insight generation (Ghobakhloo et al., 2024). IoT sensors embedded across production and logistics streams provide a shared signal on equipment condition, flow, and supply status that supports adaptive learning and early weak-signal detection (Ghaleb et al., 2020; Vilkas et al., 2024). Big-data platforms and AI analytics render these signals actionable by uncovering recurrent patterns and generating predictive and prescriptive outputs that guide attention and triage (Bag et al., 2023; Silva et al., 2021). Where uncertainty is high, generative methods expand scenario exploration by producing alternative futures and synthetic data to probe detection thresholds (Ghobakhloo et al., 2024; Silva et al., 2021).

Within seizing, technologies that enable synchronized decisions and agile resource moves are most salient. Cloud connectivity supports location-independent access to shared information and coordinated adjustment across teams and sites (Corvello et al., 2023; Roy et al., 2024). In contexts where multi-party trust, traceability, and auditability are binding, blockchain can strengthen collaboration and automate inter-organizational coordination (Ghobakhloo et al., 2025). IoT supports local control actions that stabilize flow under shifting constraints, including parameter changes and event-triggered responses at machines and cells (Zhang et al., 2021). When conventional configurations are constrained, AI methods, including generative ones, can help propose feasible alternatives for short-cycle resequencing, buffer moves, or resource reallocation (Klar et al., 2024; Ghobakhloo et al., 2024).

Within transforming, emphasis shifts to validated change and disciplined learning. Digital twins allow low-risk experimentation and pre-implementation evaluation of process, layout, or coordination changes before they are introduced into live operations (Bellavista et al., 2023; Lepore et al., 2022). AI-enabled learning systems consolidate feedback from prior episodes into improved models and decision rules, supporting operational dynamism over time (Belinski et al., 2020; Ghobakhloo et al., 2024). Generative AI can broaden the option set for design and process changes when conditions shift, accelerating the review of viable alternatives (Ghobakhloo et al., 2024). Experts also noted the importance of decentralized decision environments that allow local teams to apply insights consistently, which supports self-improvement and diffusion of new routines (Bag et al., 2023; Fosso Wamba et al., 2024).

Several technologies act as amplifiers whose salience depends on product, process, and workforce context. AR and VR primarily enhance learning, training, and guided execution, which improves sensing quality and accelerates the institutionalization of improvements during transforming (Egger and Masood, 2020; Munir et al., 2022; Roy et al., 2024). Additive manufacturing is most consequential where rapid reconfiguration, localization, or customization is required; it supports transforming by lowering changeover barriers and, in specific cases, near-term seizing through fast part availability (Iqbal et al., 2020). In inter-organizational settings where verification and shared records are pivotal, blockchain contributes to trusted coordination; in the absence of these conditions its incremental effect is smaller relative to the core data and intelligence layer and the context-specific amplifiers (Ghobakhloo et al., 2025).

In sum, Table 3 organizes technologies by how they help build capabilities rather than treating them as interchangeable tools. Shared data and analytics resources enable sensing, seizing, and transforming across settings, while other technologies matter most in specific product, process, or inter-organizational contexts. This structure stays true to the study’s theoretical intent, reflects the validated links, and shows how digitalization supports the capability system needed for antifragility.

The Delphi results suggest that antifragility in manufacturing is best understood as a system of interdependent mechanisms across sensing, seizing, and transforming. The eight capabilities that reached consensus cohere as a logic of action that is compatible with dynamic capability thinking yet differs in orientation. Rather than aiming primarily at recovery to baseline or balanced functioning, the target is improvement because of volatility, which places the results closer to the emerging view that variability can be harnessed when learning and redesign are institutionalized as routine outcomes of disruption (Ghobakhloo et al., 2025; Munoz and Zhou, 2023).

Interpreting the pattern begins with early recognition. Alertness, Adaptive Learning, and Proactive Risk Management cluster because faster and more accurate detection requires weak-signal scanning, disciplined interpretation, and pre-commitment to viable responses. This corroborates long-standing arguments in resilience and reliability research that emphasize monitoring, preparedness, and anticipation as foundational to adaptation (Mandal, 2019; Tiwari et al., 2024). The present results extend that view in two ways. First, they position recognition not only as protective but as developmental, since better sensing today should raise the quality and speed of sensing tomorrow through feedback from transforming. Second, they emphasize that risk practices are most useful when they enable action rather than cataloging threats, which aligns with proactive risk perspectives that couple assessment with option generation.

The mobilization stage is explained by Bricolage and Dynamic Collaborations. In resource-constrained settings, bricolage enables creative recombination of what is already available, while collaboration opens access to complementary resources and knowledge across partners. This pattern is consistent with supply chain and operations studies that link transparency, information sharing, and flexible coordination to performance under shocks and demand swings (Hamann-Lohmer et al., 2023). It also fits evidence from human–machine teaming where collaborative robots help reassign work and rebalance flow when variability increases (Zafar et al., 2024). The results therefore corroborate prior work on the value of coordination and improvisation, while extending it by placing these mechanisms within a loop whose explicit aim is cumulative capability gain, not just short-term continuity.

A key conceptual clarification is the separation between Operational Responsiveness and Operational Dynamism. Responsiveness adjusts execution within the current operating design, often on short cycles, through changes to parameters, schedules, or local resource allocations. Dynamism changes the design itself when conditions demand it, often on longer cycles, through reconfiguration of assets, processes, and decision rights. The literature often treats these as points on a single continuum of flexibility. The present study suggests they are different levers whose misapplication can be costly. Overreliance on Responsiveness where structural change is required leads to chronic firefighting. Initiating redesign where short-cycle control would suffice creates churn and crowding out of execution. This distinction complements dynamic capability theory, which differentiates between first-order adjustments and higher-order reconfiguration, and it aligns with our positioning of transforming as the hinge that governs renewal over time (Ghobakhloo et al., 2024).

Self-Improvement closes the loop. After-action reviews, standard-work updates, and playbook revisions ensure that lessons are retained and diffused. This is consistent with organizational learning research in operations, which links codification and reuse to performance stabilization. The contribution here is directional. Learning is not only aimed at variance reduction; it is also aimed at raising future returns under disorder by reducing detection and adjustment time and lowering the coordination cost of change in subsequent episodes (Wolf et al., 2021; Lafuente and Sallan, 2024). This helps explain why antifragility is cumulative. The benefits accrue through repeated cycles where transforming feeds back into sharper sensing and more reliable seizing.

Situating the findings within adjacent literatures shows both continuity and extension. Relative to resilience, which prioritizes recovery, the results add a feedback mechanism that targets improvement as the default outcome when volatility is persistent. Relative to viability, which seeks balance between stability and adaptability, the results tilt the balance toward growth through well-governed change. Relative to the dynamic capability tradition, the results specify microfoundations for a looped interaction in which transforming is not a terminal stage but a generator of future sensing and seizing capability. This interpretation aligns with, and extends, recent work that links capability building to technology context and collaboration while stopping short of articulating improvement because of volatility as a first-order objective.

Digitalization enters the picture as mechanism rather than catalogue. Shared data and analytics resources provide the information and intelligence base that sensing, seizing, and transforming require, which fits evidence on IoT, cloud integration, and analytics improving detection, coordination, and review (Ghobakhloo et al., 2024; Roy et al., 2024). Other technologies amplify particular parts of the loop when conditions warrant. Where multi-party coordination depends on verification and traceability, auditability tools increase the feasibility of collaboration and shorten time to action. Where redesign must be validated before physical change, simulation and visualization through digital twins and AR or VR reduce risk and speed diffusion of new routines (Li et al., 2023; Bellavista et al., 2023; Lepore et al., 2022). This role-based reading supports recent findings on technology-enabled collaboration and innovation yet moves beyond prior lists by linking technologies to specific capability effects that advance the antifragility loop.

The results also surface contingencies and risks that qualify scope. Differences in digital maturity, process coupling, product clockspeed, and network structure matter. Capabilities such as cross-functional governance, IT or OT maturity assessment, and digital resource readiness can be resource intensive, which makes sequencing important. Smaller firms can still progress by focusing on narrow, high-leverage routines that create the loop, such as weak-signal reviews, brief after-action learning cycles, and safe-to-fail experiments tied to operational goals, and by using lightweight assessments, simplified governance practices, and ecosystem partnerships for expertise and infrastructure. This reading is consistent with digital transformation studies that caution against tool-led deployments and fragmentation under resource constraints (Sony et al., 2022; Ghobakhloo and Iranmanesh, 2021). Risks follow directly. Poor data quality elevates false alarms and erodes trust in sensing. Deployments that are not anchored in capability priorities fragment effort and increase fragility. Collaboration tools can generate new dependencies if decision rights are unclear. These risks reinforce the importance of the catalyst capabilities documented earlier, which channel technology toward capability growth and maintain coherence during change.

The mechanism implies testable propositions for future work even though the study itself is conceptual and expert-based. Transforming that consistently codifies learning and revises the operating design should be associated with shorter detection and adjustment times in subsequent disruption episodes. The joint presence of Bricolage and Dynamic Collaborations should lower the cost of first actions when constraints bind, relative to either capability in isolation. Misalignment between the chosen lever of change, whether short-cycle Responsiveness or structural Dynamism, and the disturbance type should be associated with higher variance and lower performance gains from volatility. These propositions can be examined through longitudinal field designs that track capability growth and performance over repeated disturbances, through process tracing in multi-site cases that reveal decision rights and coordination patterns, and through pilot interventions that couple targeted capability changes with measurable outcomes.

Finally, the expert-based framework in Figure 2 integrates the capability set, catalyst capabilities, and enabling technologies into a single design logic. Cross-functional governance aligns initiatives with antifragility aims and reduces fragmentation across functions (Papagiannidis et al., 2023; Wang et al., 2022). Interoperability and scalability assessment reduces integration risk when new tools must coexist with legacy systems and helps maintain cohesion during change (Jimeno-Morenilla et al., 2021; Pan et al., 2021). Risk-responsive integration embeds detection and mitigation into deployment so that systems react as issues arise rather than after the fact (Riahi et al., 2021; Spieske and Birkel, 2021). IT or OT maturity assessment identifies structural gaps that otherwise become points of fragility during digitalization (Ghobakhloo and Iranmanesh, 2021; Qader et al., 2022; Ehie and Chilton, 2020). Supply chain digitalization alignment extends these logics across the network so partners share standards and signals for coordinated responses to volatility (Chi et al., 2020; Ghobakhloo et al., 2025). Digitalization strategic management links roadmaps, resources, and monitoring to antifragility aims, and digitalization resource readiness builds skills and financial flexibility for sustained implementation (Santos et al., 2023; Gupta et al., 2022; Ansari et al., 2023). Under these conditions, enabling technologies support the eight capabilities in a way that makes the sensing, seizing, and transforming loop repeatable and value creating. Over repeated cycles, firms should observe faster detection, lower cost of first actions, fewer frictions in redesign, and ultimately performance gains under volatility relative to prior baselines.

This study has elaborated on the concept of antifragility in manufacturing by developing a framework that links advanced technologies with strategic capabilities to create systems that strengthen through disruption. Theoretically, the study advances understanding of antifragility by tracing its evolution and clarifying how it differs from resilience, robustness, and viability. While those perspectives emphasize restoring or preserving performance under stress, antifragility reframes disruption as a source of improvement. This is not a semantic shift but a reorientation: disorder becomes an input for capability development rather than a threat to be contained. For manufacturing, the design objective is no longer recovery or balance but purposeful strengthening through variability. This view extends resilience and viability research by showing how firms can move beyond stability toward systematic gains in volatile environments.

The framework also contributes to dynamic capability theory by conceptualizing antifragility as a higher-order capability with distinct mechanisms. Building on the established sensing, seizing, and transforming logic, it specifies how these activities interact iteratively to create cumulative improvement. Transforming does not merely end a response episode, as it codifies lessons and reconfigures routines in ways that enhance subsequent sensing and seizing. This self-reinforcing cycle clarifies why antifragility is more than adaptability. It identifies how variability can be converted into durable advantage, and it delineates boundaries with resilience, which focuses on recovery, and viability, which balances stability and adaptability.

A further contribution lies in digital transformation research. The framework positions digitalization not as an end in itself but as a means to capability building. Many digital initiatives falter because they are tool-led and fragmented. By linking antifragility to organizational mechanisms such as cross-functional governance, digital resource readiness, and risk-responsive integration, the study shows how technologies and organizations co-evolve under turbulence. This integrated perspective connects microfoundations of capability development with the strategic role of digitalization in volatile contexts. The formulation is necessarily provisional, yet it clarifies constructs, specifies mechanisms, and sets a basis for empirical testing of scope conditions and performance effects.

Beyond theory, the framework generates practical implications. Antifragility is proposed as both a novel concept and a viable strategy for manufacturers in unpredictable environments. The central implication is directional: organizations should design recurring cycles of sensing, seizing, and transforming so that disruptions generate cumulative learning and growth. Unlike resilience, which prioritizes recovery, antifragility is an orientation toward systematic improvement and should be treated as a long-term, iterative process.

Practical feasibility depends on firm size and readiness. Supporting capabilities such as IT/OT maturity assessment, cross-functional governance, and digitalization resource readiness are resource-intensive and may appear inaccessible to smaller firms. Yet SMEs are not excluded. Progress can be achieved incrementally through simplified maturity assessments, lightweight governance that clarifies decision rights, and ecosystem partnerships that provide external expertise. By sequencing investments and focusing on feasible steps, smaller firms can gradually embed antifragility while still benefiting from its underlying logic.

Another implication concerns technology adoption. Technology should follow capability priorities rather than drive them. Firms should first identify whether their most pressing need is to strengthen sensing, seizing, or transforming, and then select tools that directly support those aims. IoT and predictive analytics can enhance sensing, while digital twins and additive manufacturing may be more relevant for transforming. A pragmatic starting point is to deploy a limited set of tools that address immediate needs, such as IoT monitoring of critical assets, basic cloud integration for visibility, or targeted AI for urgent problems. As routines mature, these foundations can be expanded with more advanced technologies such as digital twins for testing reconfigurations or AR/VR for workforce training. For SMEs, phased adoption through pilot projects with clear goals and reinvestment of demonstrated gains reduces risk and supports sustainable progress. In this way, technology decisions remain tied to capability building, avoiding fragmented investments and ensuring that digitalization contributes directly to antifragility.

Leadership attention is a critical enabling factor. Managers should assign responsibility for maintaining the sensing–seizing–transforming cycle, establish regular reviews of lessons learned, and ensure that improvements are codified and reused. Tracking simple indicators such as detection speed, adjustment time, and share of lessons incorporated into practice helps sustain focus on capability growth rather than technology deployment alone. This reinforces that antifragility is as much about organizational learning and redesign as it is about tools.

The framework also aligns with sustainability where environmental initiatives strengthen adaptability, efficiency, or responsiveness. Circular systems that reduce dependence on volatile inputs, or energy-efficiency programs that create operational slack, illustrate how sustainability and antifragility can reinforce one another. Evaluating sustainability projects through an antifragility lens helps firms prioritize initiatives that deliver both environmental and organizational benefits.

Antifragility thus becomes actionable when pursued as a program of continuous, small-scale improvements that accumulate over time. Large firms may deploy the full set of governance mechanisms and capabilities, while SMEs can progress incrementally by sequencing investments, leveraging partnerships, and focusing on routines that turn disruption into lasting learning. Across contexts, sustainable advantage derives less from owning every technology and more from aligning selected technologies with organizational routines and decision rights that make sensing, seizing, and transforming mutually reinforcing.

The study also acknowledges limitations, which point to important directions for future research. The framework remains conceptual and expert-informed rather than empirically validated. A central element, the self-reinforcing cycle linking sensing, seizing, and transforming, is theoretically reasoned and based on expert judgment rather than direct observation. Empirical testing is essential to confirm this mechanism and establish its performance effects. Longitudinal field studies could track how capabilities evolve and how firms perform under volatility, using indicators such as disruption detection speed, adjustment speed, share of lessons retained, and changes relative to pre-disruption baselines. Research designs employing panels, event windows, or difference-in-differences could strengthen causal inference. Complementary case studies can provide insight into process dynamics and decision-making within the cycle, drawing on evidence such as sensor logs, rescheduling records, and change requests. Multi-site cases across industries and firm sizes, including SMEs, could reveal how digital maturity, supply network structures, regulatory contexts, and governance models shape outcomes.

Another priority is measurement development. Future work should build constructs that distinguish antifragility from resilience and viability and capture systematic improvement through volatility. Scale development should follow established steps, including expert review, pilot testing, and confirmatory factor analysis, ideally complemented by archival or operational data to reduce common method bias. Targeted studies of specific technologies are also needed. Quasi-experimental pilots or design science interventions could test, for example, how digital twins affect transforming or how IoT and analytics enhance sensing. Mixed-methods designs that combine quantitative outcomes with qualitative narratives would provide richer evidence of how technologies contribute to capability growth. Replication across sectors and geographies should test generalizability and examine moderators such as firm size, partnerships, and governance intensity.

The data that support the findings of this study are available from the corresponding author, Dr Morteza Ghobakhloo, upon reasonable request.

While preparing this work, the authors used AI-based editing tools to proofread some parts of the manuscript. After using this tool/service, the authors reviewed and edited the content as needed and take responsibility for the publication’s content.