A Data-Driven Strategy for the Structural Health Monitoring of the Machico Cable-Stayed Bridge: A Case Study

Abstract

The management of aging infrastructure requires a paradigm shift from routine, time-based inspections to data-driven, condition-based assessment. This paper presents a novel and practical framework for this transition through an in-depth case study of the Machico Cable-Stayed Bridge, a modern structure exhibiting accelerated deterioration driven by its aggressive marine environment. The core contribution is a replicable methodology demonstrating how to leverage a unique and disparate set of existing diagnostic data—synthesizing visual inspection reports, non-destructive evaluations, and dynamic in situ measurements—to design a targeted and optimized Structural Health Monitoring (SHM) strategy. The diagnostic analysis reveals critical pathologies, including advanced corrosion and significant discrepancies between design and measured cable forces, a finding that highlights a critical gap in historical performance data. In direct response to an official mandate for continuous monitoring, this paper proposes a multi-level SHM framework where the placement and specifications of each sensor (accelerometers, strain gauges, corrosion sensors) are directly justified by the documented failure modes. This work thus provides a practical roadmap for translating forensic data into a coherent, long-term asset management strategy, bridging the critical gap between diagnostic engineering and proactive infrastructure management.

1. Introduction

The global network of civil infrastructure, particularly bridges constructed in the mid-to-late 20th century, is facing unprecedented challenges related to aging and performance degradation under increasing traffic loads and aggressive environmental conditions [1,2]. The management of these critical assets, which form the backbone of modern economies, requires a change in basic assumptions from traditional, reactive maintenance cycles to proactive, data-driven assessment strategies [3,4]. The consequences of neglecting this shift are not merely economic; they carry significant risks to public safety, as tragically demonstrated by several high-profile bridge collapses worldwide [5]. Consequently, there is an urgent and universally recognized need for more intelligent, dependable, and efficient methods for monitoring the health of these vital structures [6,7].

In response to this challenge, Structural Health Monitoring (SHM) has emerged as a key discipline in civil engineering [8]. SHM aims to provide a continuous, data-driven diagnosis of a structure’s condition through the deployment of sensor networks and the analysis of the data they produce [9,10]. This approach forms the foundation of the Digital Twin (DT) concept, where a virtual model of a physical asset is continuously updated with real-world data, transforming it from a static representation into a dynamic, predictive management tool [11,12]. Vibration-based SHM, which identifies changes in a structure’s dynamic properties (i.e., natural frequencies and mode shapes) as indicators of damage, is one of the most powerful and widely applied techniques for bridges [13,14]. However, a primary challenge in its real-world application is the need to decouple damage-induced changes from those caused by fluctuating environmental and operational conditions, such as temperature, humidity, and traffic [15,16]. This challenge is particularly acute for long-span bridges in harsh maritime environments, where wind-load and corrosion effects are dominant, as seen in studies on both cable-stayed and suspension bridges [17,18].

The Machico Cable-Stayed Bridge, a key structural asset for the urban and touristic mobility of Machico, Portugal, serves as a compelling case study for these challenges [7]. Inaugurated in 2009, its location in a Class XS3 marine environment, the most severe classification for chloride-induced corrosion under EN 206-1—has led to a state of accelerated deterioration that belies its relative youth [7,8]. This situation highlights a critical issue: modern design standards alone do not guarantee durability without a robust, long-term management strategy.

A series of recent, in-depth technical reports have provided a uniquely comprehensive diagnosis of the bridge’s condition. The key findings from these reports are synthesized in

Table 1. Synthesis of Pathologies and Condition Ratings from the VSL Inspection Report [11].

Component	Condition Rating (EC)	Key Pathologies Identified
Bearings	5 (Extremely Poor)	Advanced corrosion on all metallic parts; fractured concrete plinths under multiple bearings; corroded or entirely missing anchor bolts.
Stay-Cable System	4 (Very Poor)	Widespread water ingress and accumulation inside anchorage protection caps and tubes; severe corrosion on anchor plates, nuts, and tubes.
Concrete (Anchorage Zones)	3 (Poor)	Significant cracking and spalling on the concrete anchorage blocks for the stay cables, with insufficient concrete cover over the steel reinforcement.
Concrete (General)	3 (Poor)	Generalized micro-cracking and spalling on the surfaces of piers, towers, and the deck, with localized areas of exposed and corroded rebar.
Drainage System	4 (Very Poor)	Missing or inefficient drainage pipes lead to water runoff directly onto structural elements and exacerbating corrosion issues.
Expansion Joints	2 (Reasonable) to 3 (Poor)	Degraded neoprene, cracked transition bands, and absence of sealing, allowing further water penetration into the structure.

, painting a picture of a structure under significant stress. These combined pathologies—spanning from material-level degradation to component failure and global force redistribution—have prompted a proposed rehabilitation plan with an estimated cost of €625,000, which includes the complete replacement of the stay-cable system [19].

A critical conclusion from the official assessment by the Regional Laboratory of Civil Engineering (LREC) was the complete absence of a structured monitoring plan throughout the bridge’s service life [12]. This lack of performance data makes it impossible to track the evolution of the observed anomalies. Consequently, the LREC report concludes with a clear and urgent mandate: “it is recommended to implement a long-term behavioral monitoring program… to support decision-making on conservation and preventive/operative maintenance actions” [12]. This official recommendation highlights a crucial gap: despite the existence of a rich dataset from multiple, isolated diagnostic inspections, a coherent, long-term strategy to transform this data into actionable knowledge for asset management is missing.

In direct response to this officially identified need, this paper presents a comprehensive, data-driven strategy for the design of a tailored SHM system for the Machico Bridge. While many SHM studies focus on developing new algorithms or sensor technologies [20,21], this work contributes a practical and replicable framework. The primary contribution is demonstrating how to leverage a multi-source set of existing, disparate diagnostic data (visual, NDT, material tests) to design an effective, optimized, and targeted monitoring strategy that directly addresses the identified failure modes. This study aims to:

• Integrate and analyze data from visual inspections, non-destructive testing (NDT), and dynamic measurements to build a holistic diagnosis of the bridge’s current state.
• Use this diagnosis to identify the critical failure modes and sensitive structural parameters that must be monitored.
• Propose a practical and optimized sensor network layout, justifying the type and placement of sensors based on the identified needs and established principles of structural monitoring [22,23].
• Provide a clear roadmap for the implementation of a long-term monitoring program that directly fulfills the LREC’s official mandate, forming the basis for a future Digital Twin.

By focusing on a real-world case study with an exceptional level of preliminary data, this paper contributes a practical framework that bridges the gap between theoretical SHM research and the immediate, pressing needs of infrastructure asset managers.

2. The Machico Bridge: A Multi-Source Diagnosis of an At-Risk Structure

This chapter presents a detailed diagnosis of the Machico Cable-Stayed Bridge, synthesizing the findings from a comprehensive set of technical documents, including visual and non-destructive inspections, in situ dynamic testing, and an official technical opinion from a governmental engineering laboratory. This multi-source diagnosis forms the data-driven basis for the SHM strategy proposed in the subsequent sections.

2.1. Structural System and Environmental Context

The structure under investigation is a three-span, cable-stayed bridge with a reinforced concrete deck supported by two iconic, sail-shaped concrete towers [7]. It has a central span of 120 m and two adjacent spans of 45 m [7]. The stay-cable system is arranged in a fan-like pattern, with each tower anchoring four cables that support the deck [11]. The deck is supported on POT-type bearings at the abutments (E1, E2) and piers (P1, P2, P3), which are designed to be either fixed or unidirectional [11]. A schematic layout of the bridge is shown in Figure 1.

The bridge’s location subjects it to constant salt spray and high humidity, creating a highly corrosive environment that is the primary driver of its degradation. This aggressive exposure is the crucial factor contributing to the structure’s premature aging and has been identified as the main cause of the observed pathologies [7,12].

2.2. Visual and Non-Destructive Inspection Findings

A detailed visual inspection conducted by VSL Sistemas Portugal in March 2024 provides a granular “pathology map” of the bridge, systematically classifying the condition of each component on a scale from 0 (Optimal) to 5 (Extremely Poor). This “Estado de Conservação” (EC) scale is a standard industry metric used to rate the health of infrastructure components, where EC5 represents a failed or near-failure state requiring immediate intervention. The report’s conclusion is unequivocal: “there has been a worsening of the state of conservation of these two types of elements [bearings and stay cables] relative to the previous inspection (2019)” [11].

The key findings, summarized in Table 1, reveal a structure with critical vulnerabilities. The bearings, fundamental for the bridge’s structural articulation, are unanimously classified in the worst possible state. The stay-cable system, the primary load-bearing element, is severely compromised by water ingress. Figure 2 presents photographic evidence of some of the most critical pathologies documented in the report.

2.3. Material Degradation Analysis

The official opinion from the Regional Laboratory of Civil Engineering (LREC) provides scientific evidence that underpins the visual observations [12]. The results, one of which is reproduced in Figure 3, are alarming. The LREC concludes that, “in the generality of the five tested samples, the respective chloride content at the depth of the reinforcements exceeds the critical limits from which the de-passivation and corrosion of the steel is admitted” [12]. This finding confirms that the steel reinforcement is under active and widespread corrosive attack.

Furthermore, the LREC raises an important hypothesis regarding the cause of some pathologies. It suggests that the severe cracking in the anchorage blocks and bearing plinths may result not only from corrosion but also from “structural under-dimensioning for actions not foreseen in the project” [12]. This highlights the complexity of the problem and reinforces the need for a monitoring system capable of measuring the real-world strains and dynamic responses of these critical elements.

2.4. Diagnosis from Material Testing and Official Opinion

In March 2024, the tensile forces in the eight stay cables were measured using the VSL Vibratest system, a non-destructive testing (NDT) method based on ambient vibration analysis [13]. The method involves attaching portable accelerometers to a cable to measure its natural frequencies of vibration. The tensile force (T) is then estimated using the “vibrating string” formula, which relates tension to the n-th natural frequency (f_n), cable length (L), and linear mass (µ) (i.e., T ≈ µ (2 L f_n/n)²). While the accuracy of this common industry method can be influenced by factors such as the cable’s flexural rigidity and boundary conditions, it provides a reliable in situ force estimation²⁵. This test provided two critical insights into the bridge’s structural state, as detailed in

Table 2. Comparison of Design Forces vs. In-Situ Measured Forces for Stay Cables [13].

Stay Cable ID	Design Force (kN)	Measured Force (kN)	Ratio (Measured/Design)
R1	804	879	109%
R2	391	672	172%
F1	669	765	114%
F2	293	661	226%
F3	1183	1176	99%
F4	703	848	121%
R3	820	1307	159%
R4	1339	849	63%

.

Firstly, the results reveal major deviations from the original design. Several cables are carrying significantly more load than intended (e.g., F2 at 226%), while at least one has experienced significant stress loss (R4 at 63%). This indicates a global force redistribution within the structure that is drastically different from the design assumptions, a finding consistent with the LREC’s hypothesis of unforeseen structural actions.

Secondly, the VSL report concludes that it is impossible to know whether these discrepancies are due to initial tension errors during construction or due to performance changes over time (e.g., prestress losses, support settlement, etc.), as no initial or continuous measurements exist [13]. This lack of a dependable “time-zero” baseline is a critical knowledge gap that prevents any assessment of the bridge’s performance evolution and underscores the absolute necessity for a continuous monitoring system.

Regarding the measurement accuracy, the VSL technical report [13] provides a detailed theoretical and practical assessment. As noted in Section 3.3 of that report, the expected theoretical precision of the tension, based on an error propagation analysis, is approximately 2.3% [13]. This is further validated in Section 3.4, which compares the Vibratest method against direct hydraulic jack lift-off measurements on other bridge projects, showing practical discrepancies of no more than 2.0% [13].

This level of measurement uncertainty (approx. 2–3%) is crucial for interpreting the results in Table 2. It confirms that while minor variations (e.g., Cable F3 at 99% or R1 at 109%) might be partially influenced by measurement error, the extreme ratios observed in Cable F2 (226%) and Cable R4 (63%) are significant, real deviations. This strengthens the conclusion that these discrepancies are not measurement artifacts but rather substantial differences between the as-built state and the design intent, underscoring the lack of historical performance data.

3. A Data-Driven SHM Strategy for the Machico Bridge

Based on the comprehensive, multi-source diagnosis presented in Section 2, this section outlines a data-driven strategy for the design and implementation of a long-term Structural Health Monitoring (SHM) system for the Machico Bridge. The strategy is designed to be practical, targeted at the identified risks, and aligned with the official LREC recommendation for continuous behavioral monitoring [12].

3.1. Monitoring Objectives

The primary objectives of the proposed SHM system are derived directly from the documented pathologies and knowledge gaps. The system must be capable of:

• Establishing a Reliable Performance Baseline: A core issue identified is the lack of historical performance data, making it impossible to understand the evolution of the bridge’s structural state [13]. A primary objective, therefore, is to establish the first-ever comprehensive and continuous baseline of the bridge’s structural and dynamic behaviors, ideally following the planned rehabilitation works.
• Early-Warning for Critical Failure Modes: The system must be designed to provide early warnings of the re-initiation or progression of the specific failure modes identified, particularly corrosion at critical locations, water ingress in the stay-cable anchorages, and anomalous stress redistributions [11].
• Decoupling Environmental Effects from Structural Damage: A significant challenge in SHM is distinguishing structural changes from the natural variability caused by environmental factors like temperature [24,25]. The system must, therefore, continuously measure key environmental parameters to allow for the quantification and decoupling of these effects during data analysis [26].
• Initiative-taking Maintenance Decisions: The goal is to provide the asset owner, the Machico City Council, with quantitative, actionable data to support a transition from a reactive to a condition-based, initiative-taking maintenance strategy, optimizing the lifecycle management of the structure [27,28].

3.2. Proposed Instrumentation and Sensor Network

To achieve the monitoring objectives, a multi-level sensor network is proposed. The selection, placement, and specifications of all sensors are directly justified by the diagnostic findings (Section 2) and established best practices. The proposed layout is illustrated schematically in Figure 4.

Level 1: Global Dynamic Monitoring (Vibration)

• Technology: Triaxial accelerometers (e.g., high-sensitivity MEMS or piezoelectric type) with a suitable dynamic range and low noise floor.
• Placement: A minimum of five accelerometers is proposed: one at the top of each concrete tower (Tower 1, Tower 2) and three along the main span of the deck (at mid-span and quarter-spans). This configuration is standard for reliably identifying the dominant global vertical, lateral, and torsional modes via Operational Modal Analysis (OMA) [23,29].
• Justification and Acquisition: This network will establish the dynamic “fingerprint” of the rehabilitated bridge. Based on literature for similar concrete cable-stayed bridges (120 m main span), the first modal frequencies are expected in the 0.5–2.5 Hz range. To accurately capture these modes and their harmonics, a synchronous sampling rate of 100 Hz is proposed. Time synchronization is critical for OMA and will be achieved using a GPS timestamping or Precision Time Protocol (PTP) enabled data acquisition system (DAS).

Level 2: Critical Component Monitoring (Strain and Corrosion)

• Technology: A combination of vibrating wire (VW) strain gauges (for long-term stability), half-cell potential sensors (for corrosion activity), and embedded moisture and temperature sensors.
• Placement:

  ○

  Stay-Cable Anchorages: VW strain gauges and half-cell sensors will be installed in the concrete anchorage blocks, which were identified as highly fissured and vulnerable (EC4) [11]. Moisture sensors will be placed inside the anchorage protection caps to provide a direct early warning of water ingress.

  ○

  Bearings: VW strain gauges will be placed on the concrete plinths beneath the new bearings to monitor stress concentrations or cracking, directly addressing the “sub-dimensioning” concerns raised by the LREC [12].

• Justification and Acquisition: This level provides direct, localized data on the health of the most critical components. As these are slow-changing phenomena, a sampling rate of one reading per hour is sufficient.

Level 3: Environmental and Load Monitoring

• Technology: An integrated weather station (measuring air temperature, humidity, wind speed/direction). Additional surface-mounted temperature sensors for the concrete deck.
• Placement: The weather station will be placed at mid-span to capture representative conditions.
• Justification and Acquisition: This data is essential for implementing the data normalization techniques (see Objective 3) needed for reliable, long-term damage detection [16,24,25]. Data will be sampled concurrently with Level 2 sensors.

3.3. System Architecture and Practical Implementation

A robust SHM system must be practical to install and maintain, especially in the target XS3 environment.

• Data Acquisition and Telemetry: A central, modular Data Acquisition System (DAS) will collect data from all sensors. Given the bridge’s location, data will be transmitted wirelessly via an industrial 4G/5G modem to a secure cloud-based server for storage, processing, and visualization.
• Power: The system will be designed for autonomous operation. Power will be supplied by a solar panel system with battery backup, eliminating the need for unreliable grid power on the structure.
• Marine-Grade Protection: All external components, including sensors, cabling, and junction boxes, will be specified with a minimum IP67 rating and housed in corrosion-resistant (e.g., 316 stainless steel or marine-grade polymer) enclosures to withstand the harsh salt-spray environment.
• Maintenance: A maintenance and calibration plan will be established, mandating annual visual inspections of the SHM hardware and calibration of environmental sensors as per manufacturer guidelines.

3.4. Data Management and Analysis Framework

An effective SHM system is defined not only by its hardware but also by its framework for data management and analysis. The proposed strategy includes:

• Data Acquisition and Transmission: A central data acquisition system (DAS) will collect data from all sensors. The data will be transmitted wirelessly to a secure cloud-based server for storage and processing.
• Data Processing and Cleaning: Raw data will undergo a systematic cleaning and processing phase to remove noise, manage missing data, and integrate the different data streams (e.g., synchronizing structural data with environmental data), following best practices for SHM data handling [30].
• Feature Extraction and Baseline Modeling: After an initial period of data collection post-rehabilitation, a statistical baseline model of the “healthy” structure will be established. This involves extracting key features from the sensor signals, such as modal parameters from the accelerometers and peak strains from the strain gauges.
• Anomaly Detection and Diagnosis: The continuous stream of new data will be compared against the established baseline. Statistical process control methods or unsupervised machine learning algorithms can be employed to automatically detect statistically significant deviations, or anomalies, from the normal behaviors [16,31]. Once an anomaly is detected, further analysis, such as correlating data from different sensor types and applying damage localization algorithms, can be used to diagnose the potential cause and location of the issue [32,33].

This structured framework ensures that the vast amount of data collected is transformed into concise, actionable information for the bridge managers (Figure 5).

4. Discussion

This chapter interprets the proposed data-driven strategy presented in Section 3, connecting the designed monitoring solution to the real-world structural issues diagnosed in Section 2. It discusses the practical implications for the management of the Machico Bridge, its alignment with the state-of-the-art in SHM, and the limitations of the current study.

4.1. From Diagnosis to Strategy: A Data-Driven Pathway

The core strength of the proposed SHM strategy lies in its foundation on a rich set of diagnostic data. Rather than a generic, one-size-fits-all approach, every aspect of the monitoring plan is a direct response to a documented pathology or an officially identified knowledge gap. The classification of the bearings as being in an “Extremely Poor” (EC5) condition, with fractured concrete plinths, directly justifies the proposal of strain gauges in these specific locations to monitor local stress concentrations and the performance of future repairs [11]. Similarly, the finding of widespread water ingress in the stay-cable anchorages is the primary driver for proposing embedded moisture and corrosion sensors, providing a much-needed early warning system for the main cause of degradation [11]. This data-driven pathway therefore serves as the critical ‘connection’ between the before state (a structure diagnosed with forensics but lacking performance history) and the after state (a fully data-enabled asset prepared for predictive management).

Furthermore, the data from the VSL Vibratest report, which revealed significant uncertainty regarding the as-built cable forces, underscores the necessity of a global dynamic monitoring system [13]. A network of accelerometers capable of performing Operational Modal Analysis (OMA) is the only way to establish a reliable dynamic baseline for the structure. This baseline is critical, as changes in dynamic properties are sensitive indicators of significant structural changes, including the effects of corrosion on the overall stiffness [34]. This data-driven pathway ensures that the proposed SHM system is not merely a collection of sensors, but a targeted diagnostic tool designed to observe the most critical and uncertain aspects of the bridge’s behaviors.

4.2. Implications for the Management and Rehabilitation of the Machico Bridge

The implementation of the proposed SHM system has profound implications for the lifecycle management of the structure, particularly concerning the planned €625,000 rehabilitation [19]. The system offers value in two distinct phases:

• Post-Rehabilitation Verification: Once the rehabilitation works are complete, the SHM system will serve as an essential quality control and performance verification tool. It will monitor the behaviors of the new components (especially the new stay cables and bearings) under operational and environmental loads, ensuring they are performing as designed. This process will establish the new, “healthy” baseline, a critical step often overlooked in repair projects, and is consistent with methodologies used to assess bridges before and after strengthening works [35].
• Transition to Initiative-taking, Condition-Based Maintenance: The goal of SHM is to facilitate a shift from reactive, time-based maintenance to an initiative-taking, condition-based philosophy [22,27]. The continuous data stream will allow the asset owner to track the slow progression of degradation, quantify the rate of change, and make informed decisions about when and where to intervene. This initiative-taking approach is widely recognized as more cost-effective over the lifecycle of an asset, helping to prevent minor issues from escalating into the need for major, costly interventions like the one currently required [31,36].

4.3. Alignment with the State-of-the-Art and Future Work

The proposed multi-level monitoring strategy (global dynamics, local critical components, and environmental factors) is well-aligned with comprehensive SHM frameworks presented in the current state-of-the-art literature [3,37]. The use of OMA as the core of the global monitoring level is a well-established and powerful technique for bridges of this type [23,38]. The explicit inclusion of environmental sensors to decouple temperature effects from structural changes addresses one of the most significant challenges in long-term SHM and is a focus of much current research [16,24,25].

This work lays the foundation for several future steps that would further enhance the management of the bridge:

• Implementation and Digital Twin Integration: The physical implementation of the sensor network is the logical next step. The data stream generated would serve as the foundation for a true Digital Twin of the Machico Bridge, enabling predictive analysis and scenario simulation [14,15,39].
• Advanced Data Analytics: Once enough data is collected, advanced data-driven methods, such as unsupervised machine learning algorithms, could be applied to automatically detect anomalies and identify damage patterns that might be missed by traditional analysis [16,31].
• Refined Damage Localization: The data from the accelerometer network could be used with advanced algorithms, such as those based on Modal Strain Energy, to not only detect but also localize potential damage along the bridge deck, providing even more targeted information for inspectors [26].

A note must also be made on the use of advanced data analytics, such as the machine learning (ML) algorithms mentioned in Section 3.3. We acknowledge the “black box” or interpretability challenge of ML in engineering applications. Our proposed framework prioritizes interpretability; ML algorithms (e.g., unsupervised clustering) would be used primarily for anomaly detection to identify statistically significant deviations from the established baseline. The crucial step of damage diagnosis would then be performed by correlating these anomalies with data from other sensors (e.g., an anomaly in OMA data correlated with a moisture sensor alarm) and validated by qualified engineers, thereby using ML as a powerful screening tool rather than an opaque decision-maker [40].

4.4. Limitations and Future Work

It is important to acknowledge the limitations of this study. The work presents a comprehensive strategy for SHM, not its physical implementation. The true effectiveness and challenges of the system can only be fully assessed after its installation and commissioning.

Furthermore, this study deliberately forgoes the creation of a Finite Element Model in favor of a strategy based purely on the available diagnostic data. While this demonstrates the power of a data-driven approach, a calibrated FEM would be a valuable future addition. Such a model, updated with data from the proposed SHM system, would allow for a more refined optimal sensor placement and enable quantitative damage prediction through numerical simulations [41,42,43].

It is also important to acknowledge the inherent challenges of the advanced data analytics proposed. As the system collects long-term data, it will be essential to manage large data streams and apply robust probabilistic models capable of handling complex “mixed skewness” characteristics from multiple sources (e.g., traffic, environment). As noted in the recent literature, the computational cost and parameter stability of such advanced models represent a significant challenge and an active area of research [44].

A final limitation, which represents a clear avenue for future work, is the deliberate exclusion of traffic load monitoring from primary SHM design. While traffic is a major load source, the documented critical pathologies (corrosion, anchor failures) were prioritized. A complete Digital Twin would benefit from the future expansion of this system. This could be conceptualized as a future “Level 2 upgrade”, integrating either Weigh-in-Motion (WIM) sensors or an array of dynamic strain gauges on the deck. This would not only quantify operational loads but also allow for linking the observed modal trends (from the Level 1 accelerometers) to probable traffic effects, thus closing the conceptual loop and demonstrating a full awareness of operational influences.

5. Conclusions

This study was motivated by the documented accelerated deterioration of the Machico Cable-Stayed Bridge and a formal recommendation from the Regional Laboratory of Civil Engineering (LREC) for the implementation of a continuous monitoring program [12]. This paper has presented a comprehensive, data-driven strategy for the design of a tailored Structural Health Monitoring (SHM) system, leveraging a rich set of existing diagnostic data. The primary findings and contributions of this work are summarized as follows:

• A multi-source diagnosis is key to a targeted strategy. By integrating and analyzing data from visual inspections, in situ dynamic measurements, and material tests, a detailed and robust diagnosis of the bridge’s condition was established [11,12,13]. This diagnosis revealed critical pathologies related to chloride-induced corrosion and significant uncertainties regarding the as-built state of the structure, particularly the tensile forces in the stay cables.
• A data-driven SHM strategy provides a practical and effective solution. Based on the diagnostic findings, a practical and targeted SHM strategy was developed. The proposed sensor network is not generic; it is specifically designed to monitor the identified failure modes, such as water ingress at anchorages and stress concentrations at the critically rated bearing locations. This approach demonstrates the paper’s key contribution: a replicable framework for translating existing, high-quality inspection data directly into an intelligent and optimized monitoring plan.
• An actionable roadmap for modern asset management has been provided. This study delivers a clear and scientifically backed roadmap for the asset owner to implement a modern monitoring system. This system will establish the first-ever reliable performance baseline for the bridge, enable the early detection of future damage, and provide the quantitative data needed for a cost-effective, condition-based maintenance program, directly fulfilling the official mandate from the LREC [12].

In conclusion, this case study demonstrates the immense value of leveraging comprehensive diagnostic data to design intelligent and efficient SHM systems. It serves as a practical example of how to transition aging infrastructure from a reactive maintenance cycle to an initiative-taking, data-driven management philosophy. The implementation of the proposed strategy will not only address the immediate structural concerns of the Machico Bridge but also establish a framework for ensuring its long-term safety, resilience, and sustainability.