The maximum information possible is being extracted from the Pont de Normandie for real-time and long-term evaluation
The three crossings are benefitting from the comparatively new development in structural health monitoring that combines sensor fusion, big data and web-based IoT technologies. Positioned at the mouth of the Seine Estuary, the port of Le Havre is the fifth largest port in Northern Europe and crucial to the French economy. It is the country’s leading container port for French foreign trade, vehicle import and export, and is the worldwide leader for wine and spirits. Inland communications have long been important, with a crossing of the Seine necessary for unhindered access to Paris and beyond.
The strategic highway network in this region includes the three key bridges that comprise its two Seine crossings. The oldest of these is the Pont de Tancarville, a single span suspension bridge located 28km to the east of the city. Opened in 1959, this 1,420m-long bridge with a central span of 608m carries Paris-bound traffic on the N182 and A131 highways. A second Seine crossing was added in 1995, much closer to the port of Le Havre, carrying the A29 highway and providing a significantly more direct connection to the city of Caen, the southwest and Brittany. The Pont de Normandie is a cable-stayed bridge with a total length of 2,143m and a single central span of 856m; both bridges held world records when they first opened to traffic. Two kilometres to the north, on the approach to the Pont de Normandie, is the location of the third of the major bridges in the vicinity; the Pont sur le Grand Canal du Havre, a rigid steel-framed bridge traversing one of the main port channels.
In 2018 the operator of these three bridges, the Chambre de Commerce et d’Industrie Territoriale Seine Estuaire, sought assistance from James Fisher Testing Services and Cowi in designing and commissioning an upgraded structural health monitoring solution, as the existing systems approached the end of their operational lives.
As part of the refurbishment project, the team’s first task was to review the current monitoring installations. Based on the findings, they were able to configure a comprehensive solution that would deliver state-of-the-art functionality based on the latest big data, sensor fusion and internet-of-things technologies. This major refurbishment project is the first on which the team has partnered, and it combines significant synergies between the two. Namely, JFTS’ strengths as a structural health monitoring innovator, technology developer and system integrator; and Cowi’s expertise in data analytics and sensor requirements, as well as capability in using the data collected by the system to provide accurate performance modelling and insights on future structural condition.
The project involves upgrading the various existing monitoring systems that have come to the end of service life. The upgrade comprises an initial review of the existing systems, their refurbishment, and the installation of selected additional sensors. The goal is to have quantitative information to characterise both the environmental conditions at the location of the three bridges including wind conditions, temperatures and rain intensities, and the structural responses that these and the bridges’ operational loadings create.
To maximise insights into the performance of major infrastructure such as strategic highway bridges — while allowing for advanced analytical operations that might extend to the prediction of future condition using concepts such as machine learning — a large amount of real-time data will be generated from a multiplicity of sensors. For example, in the case of the new system being fitted to the Seine Estuary bridges, displacements are monitored permanently by a system of strategically-located GPS antennae. The strain of selected components is also measured, as well as cable vibration levels, to mention just some of the performance measurements continuously being recorded. This large amount of data provides the basis to characterise abnormal structural behaviours, and to inform a proactive approach while managing the bridges.
The new structural health monitoring solution provided by JFTS will be based on the company’s Bridgewatch solution, which enables automated, intelligent structural health and performance insights to inform operational decision support. The system uses real-time data acquisition and advanced processing to improve safety, manage traffic flow and minimise bridge closures due to unscheduled maintenance.
Analytical processing of recorded data in order to detect longer-term trends and predict future performance, however, requires the use of web-based, big data technologies. Many of these are readily available for the latest generation of structural health monitoring systems, proven in sectors as contrasting as online gaming, video-on-demand services, international banking and automated share market trading. Similarly, the analysis requirements of such state-of-the-art systems are able to harness the type of parallel processing approaches used in sectors such as automotive and aerospace for large, highly non-linear analysis operations.
In order to manage and leverage the advantages of big data, Bridgewatch operates on JFTS’ Smart Asset Management System. The Sams technology stack includes the operating system and related support programs and all runtime environments necessary to support the application, as well as database warehousing software and utilities for version control.
This approach provides one of the key innovations of the refurbished monitoring systems on the Seine Estuary bridges. While some of the new sensors that JFTS is specifying will better characterise the structural responses — for example, new inclinometers and strain gauges — the main innovation relates to the combination of data from all of the different monitoring systems into a single centralised software platform. This will allow correlation of data from different sensing systems, from weather stations to GPS, and hence maximise the information that can be extracted from the data, both immediately in real time as well as for longer-term trend evaluation and predictive analysis.
The Pont de Normandie’s weather station (left) has been upgraded (centre)
Rather than simply embark on a wholesale replacement of the installed monitoring hardware, JFTS is endeavouring to extract the maximum possible value from existing serviceable equipment, hence minimising the cost of the system upgrade. A significant number of sensors from previous installation phases — in some cases even from the original construction — are being re-used following a strict performance verification process. In other cases, hardware is being replaced or augmented in order to ensure complete and effective integration with the new Sams and Bridgewatch system.
The approach to system and component testing is rigorous; all sensors are tested in the factory prior to any installation works. Once the sensors are installed, including those units carried over from previous installations which have been tested and verified in situ, the full system will be subjected to commissioning tests. This process will be used to verify that the complete measurement chain works as planned, from the sensor attached to the structure to the web-based interface that displays and processes the data.
The installation of new and refurbished sensors is an ongoing process
The installation of the Sams/Bridgewatch system and associated new and refurbished sensor hardware is ongoing and the software platform to manage all data is being customised to be ready early this year. Once initial testing has been completed over the coming months, the system will provide complete continuity with the previous legacy monitoring data reports, while also enabling the creation of an accessible database of information and structural monitoring for future analytical use.
The combination of sensor fusion, big data and web-based IoT technologies is a comparatively new development in the world of structural health monitoring and offers the prospect of new levels of insight into structural health and proactive management of operations. These systems enable efficient day-to-day monitoring operations while also facilitating large-scale advanced analytical processes that can yield insights that would otherwise not be discernible. In this way, advanced analytical methods can be used to predict the path of future performance deterioration, thus enabling proactive maintenance and operational management.
But while the concepts are already well proven in other industries, they bring some clear challenges too, not least in terms of the management and effective manipulation of the vast quantity of data generated. It is also the case that automated analysis can never fully substitute for human intervention and interpretation; an area in which Cowi will continue to support the Chambre de Commerce et d’Industrie Territoriale Seine Estuaire. At its very best, structural health monitoring and effective infrastructure management thus requires the combination of these state-of-the art technologies with more conventional approaches — including that of regular visual inspections.
Matthew Anderson is head of bridges and structures at James Fisher Testing Services. Anthony Smith is an independent consultant
Machine learning and predicting future performance
The Sams/Bridgewatch system is configured to monitor a range of different bridges around the world. The system harnesses big data technologies and flexible, web-accessible parallel computing resources, to enable highly sophisticated forms of structural health monitoring to be carried out.
For example, the system can be configured to automatically execute sophisticated analysis operations based on a fixed schedule, or according to pre-set data triggers such as threshold exceedances. These can be either individual sensor inputs such as a given load, temperature, displacement or wind speed for a given direction, or a prescribed combination of such inputs or derived parameters such as bending stress. Alternatively, analyses may be triggered on demand or from another analytical process.
Machine-learning routines can be configured within Bridgewatch to predict future performance trends based on the complex data sets, which range from displacements to traffic loadings, temperatures and wind speeds. In a highly complex system such as a major suspension bridge, assessing subtle trends in the vast quantity of multi-sensor data generated can be extremely challenging. The use of deep-learning neural network methods can enable multiple layers of non-linear information processing for unsupervised self-learning performance based on large amounts of sensor-derived data. Advanced analysis routines can be used to automatically calculate trends in the expected future performance deterioration of structural details of concern.
Any divergence between the predicted performance and real-time measured data, or an increase in the predicted level of structural deterioration, could be used as a trigger for further analysis and investigation. One such analysis operation that might be used on this through-life data is that of stochastic subspace identification.
For a welded structure, for example, there will be a known overall stiffness on completion. If a fracture occurs in part of the structure the stiffness will be changed, but it may not be possible to spot this in the complexity of recorded data, due to other factors such as prevailing temperatures, wind speeds and traffic loadings.
Stochastic subspace identification enables the numerical manipulation of systems where not all variables are known, in order to help identify and isolate the causes of deterioration based on the live loadings applied.
Early detection of root causes for even comparatively incipient problems should thus be identifiable, enabling early rectification or even preventative action to be taken in a planned way, avoiding the costs and disruption of the more intensive interventions that might otherwise be required at a later stage of deterioration.