Instrumentation and Measurement Methods for Industry 4.0 and IoT

The current industrial transformation is changing the way production systems are designed, manufactured, monitored, controlled, and maintained. Compared to traditional automation, this transformation is characterized by pervasive connectivity, cyber-physical integration, distributed intelligence, and the seamless exchange of information between physical assets and digital services. In this scenario, the Internet of Things (IoT), artificial intelligence, cloud and edge computing, digital twins, advanced sensors, and data analytics are no longer independent technologies. They are progressively becoming elements of a single technological ecosystem where measurement data represents the essential interface between the physical and digital worlds [1]. The COVID-19 pandemic has accelerated several aspects of this transition, making remote working, telemedicine, online services, distributed monitoring, and digital communication essential tools for business continuity and social resilience. This acceleration has highlighted the need for always-connected, remotely accessible, and data-driven systems. Such systems require the ability to observe, measure, transmit, and interpret physical phenomena remotely. For this reason, the development of new sensors, acquisition techniques, data processing methods, and distributed measurement architectures has become increasingly important. This transition has profound consequences for metrology. Industry 4.0 can only create value if decisions are based on reliable, traceable, and timely information. Smart factories, in-line inspection, predictive maintenance, remote monitoring, and real-time control require measurement systems capable of operating in complex environments, often under variable operating conditions and with limited human supervision. The concept of Metrology 4.0 summarizes this evolution, emphasizing the digitalization of metrological services, the integration of measuring instruments into production systems, and the need for real-time calibration, process monitoring, traceability, and quality control [2]. From this perspective, metrology is not simply a production support activity, but an enabling infrastructure for smart, sustainable, and resilient manufacturing. The adoption of Metrology 4.0 requires qualified personnel, adequate infrastructure, management commitment, awareness of Industry 4.0 technologies, investment capacity, flexible organizational culture, and appropriate standards and norms [2]. Recent work on smart coordinate measuring machines confirms this trend, demonstrating that dimensional metrology is moving from isolated inspection tools towards connected, multi-sensor, and uncertainty-aware components of cyber-physical manufacturing systems [3]. Similar needs are emerging in additive manufacturing, where smart metrology combines in situ sensing and machine learning-based analytics to support defect detection, process optimization, predictive quality monitoring, and future closed-loop control [4]. Modern industrial and IoT scenarios require measurement systems that are not only accurate but also connected, scalable, interoperable, and able to operate under complex and variable conditions. Sensors, embedded electronics, wireless networks, edge and cloud computing, artificial intelligence, and advanced signal processing techniques are increasingly integrated into measurement chains. This convergence opens up new opportunities for real-time monitoring, diagnostics, localization, tracking, virtual measurements, human-centric applications, and smart devices. At the same time, it raises challenges related to calibration, uncertainty assessment, reliability, cybersecurity, energy efficiency, data governance, and long-term stability.

Digital twins represent one of the most prominent examples of this convergence. Their reliability depends on the continuous alignment between the physical system and its virtual counterpart. This alignment requires calibrated sensors, validated models, synchronization strategies, and methods for managing measurement uncertainty. Recent studies have highlighted the importance of calibration methods based on computer vision, event-based sensing, and neuromorphic vision to improve the accuracy of digital twins in cyber-physical manufacturing systems [5]. Verification and validation are equally important, as a digital twin can only support decisions when its credibility, fidelity, and domain validity are clearly assessed [6].

Artificial intelligence has become one of the most important tools for extracting actionable insights from industrial data. AI can support anomaly detection, fault diagnosis, production planning, quality control, adaptive monitoring, and maintenance optimization [7]. However, its adoption in Industry 4.0 is still limited by issues related to system integration, heterogeneous data, workforce factors, and AI reliability [1]. From a measurement perspective, these issues must also be considered as metrological problems. A forecast cannot be fully useful if its uncertainty is unknown, if its outcome cannot be explained, or if the data acquisition process is not reproducible.

Predictive maintenance is another central topic of Industry 4.0. In [8], for example, signals acquired by sensors are processed, and deep learning architectures are used for fault classification, while uncertainty quantification and SHAP-based interpretability provide confidence and interpretability. This approach is significant because it combines data-driven diagnosis with metrological awareness: the classification result is not treated simply as a black-box output, but as information whose reliability and physical meaning must be assessed. Other studies confirm the importance of scalable architectures for predictive maintenance. Edge-cloud and edge-fog-cloud approaches can reduce latency, distribute computational load, and support real-time decision making in resource-constrained environments [9,10]. These frameworks demonstrate that predictive maintenance is not only an algorithmic problem, but also a measurement system problem involving sensor placement, data acquisition, heterogeneous data streams, model implementation, and feedback to operators.

Security is also part of this picture. Cybersecurity, privacy, access control, and transparency in decision-making are closely related to measurement integrity. Recent studies on explainable AI for IoT cybersecurity emphasize the importance of transparent, auditable, and adaptive methods for anomaly detection and decision support in large-scale connected infrastructures [11]. This is particularly relevant when measurement data is used to guide safety-critical or economically relevant decisions.

In human- and society-centric applications, this new concept of Metrology 4.0 is emerging even more rapidly. Recent human–computer interaction systems show a development towards radar, vision, haptic, wearable, and wireless sensor technologies capable of obtaining robust information about users, environments, and activities [12]. In these systems, sensor fusion is essential because no single modality can provide complete and reliable information under all operating conditions. Healthcare and rehabilitation represent particularly important application fields. IoT and Internet of Medical Things platforms enable remote patient monitoring, wearable sensing, telemedicine, personalized care, and smart rehabilitation, while raising issues of privacy, interoperability, implementation costs, and cybersecurity [13]. Objective measurements can improve diagnosis, treatment planning, patient follow-up, and rehabilitation outcome assessment, but clinical applications require high reliability, usability, and interpretability. Antonacci et al. focused on assessing shoulder muscle strength, comparing portable dynamometers with a reference load cell in a rigid serial configuration [14]. This example demonstrates that even seemingly simple clinical measurements require attention to reference systems, mechanical setup, signal processing, and concordance analysis. Wearable and rehabilitation-oriented instrumentation further confirms the relevance of integrated sensing: an instrumented ankle-foot orthosis integrating strain gauges, force-sensing elements, an inertial measurement unit, and Bluetooth Low Energy communication can provide objective information on the gait cycle for rehabilitation [15].

Agriculture and agri-food systems provide another example of metrology’s expansion beyond traditional industrial applications. IoT, machine learning, and blockchain-based traceability can support crop monitoring, irrigation management, resource optimization, food safety, and fraud prevention [16,17]. Precision agriculture relies on distributed measurements of temperature, humidity, soil moisture, crop health, and environmental conditions. At the same time, agri-food supply chains require reliable data on production, storage, transportation, and quality [18]. The effectiveness of these digital solutions depends on the reliability and traceability of the underlying measurements.

In structural health monitoring, reproducibility and transferability are equally crucial. Chiominto et al. studied the reproducibility of impact area classification in aerospace structures using piezoelectric sensors, Lamb waves, signal processing, and machine learning models [19]. The study highlights that the performance of a classification procedure depends on feature extraction, acquisition conditions, training data, model selection, and impact point distribution. Similar issues arise in pipeline monitoring, where hydrophone-based acoustic sensing and machine learning have been investigated for robust leak detection and localization in the presence of variable sensor geometries and placements [20]. These examples confirm that machine learning-based measurement systems need to be validated not only for accuracy in a single experiment, but also for robustness, reproducibility, and transferability across devices, laboratories, and operating conditions.

Overall, the literature highlights a fundamental trend: measurement systems are becoming connected, intelligent, and application-oriented. The instrument is no longer an isolated device; it is part of a distributed architecture that includes sensing elements, integrated electronics, communication protocols, data storage, algorithms, user interfaces, and decision-making procedures. Consequently, measurement system design must consider not only sensitivity, resolution, and calibration, but also interoperability, latency, energy consumption, data security, model interpretability, and uncertainty propagation throughout the entire chain. Future research into instrumentation and measurement methods for Industry 4.0 and the IoT will therefore require a multidisciplinary approach. Sensors and acquisition systems must be designed in conjunction with data processing algorithms, physical models, validation procedures, and security mechanisms. Artificial intelligence must be combined with domain knowledge and metrological principles to obtain reliable and interpretable results. Finally, standardization, reproducible datasets, benchmarking procedures, and interlaboratory comparisons will be crucial for transferring innovative solutions from research prototypes to industrial, biomedical, agricultural, and smart environment applications. The future of Industry 4.0 and the IoT depends not only on the availability of data but also on the availability of reliable measurements. Data becomes useful only when its origin, quality, uncertainty, and meaning are known.