Predictive Maintenance (PdM) combined with Structural Health Monitoring (SHM) is revolutionizing the management of critical infrastructure such as bridges, dams, and pipelines, to face challenges like ensuring longevity and safety while adapting to climate change. These advanced technologies predict maintenance needs, reducing costly unscheduled downtime and enhancing the durability of vital structures. This article explores the application of PdM and SHM in infrastructure, detailing their operational mechanisms, benefits, and strategies for successful deployment.
Predictive Maintenance (PdM) is a proactive maintenance strategy that uses data analysis to predict when an equipment failure might occur, so that maintenance can be planned based on the actual condition of the equipment. This approach helps to prevent unexpected equipment failures and reduce unnecessary preventive maintenance activities.
PdM involves the use of various types of sensors to collect data on equipment condition. This data can include vibration, temperature, pressure, and other types of information that can indicate potential problems. Advanced analytics and algorithms are then used to identify patterns in the data that may signify a future failure.
Early stages: Predictive maintenance has been around since the early 1990s. Initially, its application was limited due to the lack of sophisticated sensors and computing power. Traditional predictive maintenance was industry-specific, focusing on machines with complex, custom-built models.
Advent of IoT and cloud computing: The advent of the Internet of Things (IoT), cloud computing, data analytics, and machine learning has significantly expanded the practicality and applicability of PdM. The IoT has enabled the collection of vast amounts of data from various sensors installed on infrastructure. Cloud computing has provided the infrastructure to store and process this data.
Role of data analytics and machine learning: Data analytics and machine learning have revolutionized predictive maintenance, transitioning it from reactive to proactive approaches. By analysing vast data sets and patterns, these technologies enable early detection of equipment issues, optimise maintenance, enhance reliability, and extend asset lifespan. Machine learning algorithms, after undergoing rigorous training on historical data encompassing sensor readings, maintenance actions, and past failures, evolve into virtual experts capable of recognizing precursors to faults.
Structural Health Monitoring (SHM) is a process used to identify damage in a structure. SHM systems can provide real-time information about the structural integrity of bridges, buildings, and other structures. This information can be used to make informed decisions about necessary repairs and maintenance, improving safety and operational efficiency.
SHM involves the use of sensors to monitor a structure over time. The data collected from these sensors is then analysed to identify any changes that may indicate damage or structural degradation. This can include changes in the material properties, geometry, and boundary conditions of the structure.
Predictive Maintenance (PdM) and Structural Health Monitoring (SHM) are like the doctors of our infrastructure. They keep a close eye on the health of our buildings, bridges, and more, much like a doctor would monitor a patient’s vital signs.
Here’s how they work together:
Here are a few examples of how Predictive Maintenance and Structural Health Monitoring can be implemented in various types of infrastructure.
Bridges are vital components of transportation networks, and their failure can have severe consequences. PdM and SHM can significantly enhance the maintenance of bridges. Sensors such as accelerometers and strain gauges can monitor vibrations and stress levels, providing real-time data on the bridge’s health. For example, accelerometers can detect unusual vibrations indicating structural weaknesses, while strain gauges measure the stress on bridge components, allowing for timely repairs.
Example: The Golden Gate Bridge in San Francisco uses a comprehensive Structural Health Monitoring (SHM) system with over 300 sensors to monitor wind speed, temperature, and stress on the bridge’s components. This system helps engineers detect potential issues early and plan maintenance activities effectively.
Dams are crucial for water supply, flood control, and hydroelectric power. PdM and SHM can help ensure their structural integrity. Sensors can monitor water pressure, seepage, and structural deformations. For instance, piezometers measure water pressure within the dam, and crack meters track the growth of fissures. By analysing this data, potential issues like leaks or structural weaknesses can be detected early, preventing catastrophic failures.
Example: The Hoover Dam employs an extensive network of over 500 sensors to monitor structural integrity, measuring factors such as water pressure, seepage, and concrete temperature. This data is critical for analysing and predicting potential problems, ensuring the dam’s safety and functionality.
Pipelines transport essential resources like water, oil, and gas. PdM and SHM can prevent failures that lead to environmental disasters and significant economic losses. Sensors can monitor pressure, flow rate, and temperature. Advanced techniques such as acoustic emission monitoring can detect leaks by identifying sound waves generated by escaping fluids. This allows for precise and timely interventions, ensuring the pipeline’s integrity.
Example: The Trans-Alaska Pipeline utilizes Predictive Maintenance (PdM) to monitor the flow and pressure of oil transported through the pipeline. Acoustic sensors detect leaks by identifying sound waves generated by escaping oil, helping to prevent environmental disasters and maintain the pipeline’s efficiency.
The foundation of an effective PdM and SHM solution is robust data collection. Sensors placed on infrastructure gather real-time data on various parameters. This data is combined with historical maintenance records and environmental information. Ensuring the accuracy and reliability of this data is crucial for developing predictive models.
Example: In a wind turbine, various sensors are installed to monitor its health. These include vibration sensors, temperature sensors, strain sensors or acoustic sensors. The data collected from these sensors is then integrated with other relevant information like historical maintenance (when and what maintenance was performed) records and environmental information (wind speed, humidity and temperature).
Data scientists use machine learning algorithms to analyse the collected data and develop predictive models. These models can identify patterns that indicate potential failures. For example, a model might predict the likelihood of a pipeline leak based on pressure fluctuations and temperature changes. By continuously refining these models with new data, their accuracy improves over time.
Example: Data scientists might use a machine learning algorithm, such as a Random Forest or a Neural Network, to analyse the collected data. These algorithms can identify complex patterns in the data that might be difficult for a human to spot.
For instance, the model might find that a specific pattern of vibration and temperature increase in a machine is often followed by a failure a few days later. This pattern would then be used to predict future failures.
Once developed, predictive models are deployed to monitor infrastructure in real-time. These models can be run on edge devices for immediate diagnostics or in the cloud for more comprehensive analysis. Continuous improvement is essential; as more data is collected, the models are refined to enhance their predictive capabilities.
The adoption of PdM and SHM in infrastructure offers numerous benefits, including:
Organizations may face different challenges when implementing PdM and SHM. These challenges include:
PdM and SHM are revolutionizing the way we manage and maintain our infrastructure. By harnessing the power of data and sophisticated analytics, these systems allow us to proactively care for vital structures like bridges, dams, and pipelines, enhancing their longevity and ensuring their safety. As we further tap into the capabilities of the Internet of Things (IoT), cloud computing, and machine learning, PdM and SHM are set to become essential tools in preserving the infrastructure that underpins our contemporary world.
Simultaneously, they are reshaping our approach to infrastructure maintenance. Through the collection and analysis of data, we’re able to anticipate and avert failures, guaranteeing the smooth and safe operation of crucial structures. As technological progress continues, the reach and impact of these systems will only grow, leading us towards a future of increasingly intelligent and efficient maintenance practices.
Sixense, as part of its predictive maintenance and SHM offering, is involved in part of the cycle, implementing permanent sensors, spot checks, digital solutions and data analysis.
Sixense offers a complete range of services: SHM+, with the aim of maintaining and extending the lifespan of infrastructure and detecting and predicting their failures. Sixense monitors both the structure and its environment (use, weather, etc.) by integrating various measurements, often automatic, including by satellite. The SHM+ offer comprises several solutions, including EverScan, EverSense®, Beyond Monitoring and Beyond Asset. It is divided into three parts to ensure the safety and longevity of your structures:
Diagnosis and design: Define instrumentation objectives and design the monitoring system with:
Continuous monitoring: deploying a smart and robust monitoring system with:
Decision support: interpret the data, assess the situation and support predictive maintenance with:
