Conference PaperPDF Available

The road to sensor-driven cloud-based infrastructure management

Conference Paper

The road to sensor-driven cloud-based infrastructure management

Abstract and Figures

Today, the accelerated degradation of many concrete structures poses a major challenge for the proper maintenance of the transport infrastructure. Therefore, inspection and maintenance operations constitute an important part of the recurrent costs of infrastructure. Furthermore, the increasing migration of population to urban areas has made sustainable development an imperative need. This need has become a driving force for innovation and new challenges such as the concept of smart cities and infrastructure. The successful utilization of newly available technologies will enable a whole new range of possibilities such as sensor driven cloud-based strategies for infrastructure management, which will promote an upgrade of the current infrastructure network to a new generation of safer, more efficient and more sustainable smart infrastructure: the infrastructure 2.0. The aim of this paper is to review the state-of-the-art of the different key technologies comprising a smart monitoring system, focusing on the aspects that are required to ensure a successful implementation of such system. The main result of the study is a scientific roadmap that can serve as a guide for traffic administrations and academic institutions in their task to develop and create a new infrastructure management strategy based on emerging technologies and innovative processes.
Content may be subject to copyright.
SynerCrete’18 International Conference on Interdisciplinary Approaches
for Cement-based Materials and Structural Concrete
24-26 October 2018, Funchal, Madeira Island, Portugal
THE ROAD TO SENSOR-DRIVEN CLOUD-BASED
INFRASTRUCTURE MANAGEMENT
Carlos G. Berrocal (1,2), Ignasi Fernandez (1), Rasmus Rempling (1,3)
(1) Chalmers University of Technology, Gothenburg, Sweden
(2) Thomas Concrete Group AB, Gothenburg, Sweden
(3) NCC AB, Gothenburg, Sweden
Abstract
Today, the accelerated degradation of many concrete structures poses a major challenge for
the proper maintenance of the transport infrastructure. Therefore, inspection and maintenance
operations constitute an important part of the recurrent costs of infrastructure. Furthermore,
the increasing migration of population to urban areas has made sustainable development an
imperative need. This need has become a driving force for innovation and new challenges
such as the concept of smart cities and infrastructure. The successful utilization of newly
available technologies will enable a whole new range of possibilities such as sensor driven
cloud-based strategies for infrastructure management, which will promote an upgrade of the
current infrastructure network to a new generation of safer, more efficient and more
sustainable smart infrastructure: the infrastructure 2.0. The aim of this paper is to review the
state-of-the-art of the different key technologies comprising a smart monitoring system,
focusing on the aspects that are required to ensure a successful implementation of such
system. The main result of the study is a scientific roadmap that can serve as a guide for
traffic administrations and academic institutions in their task to develop and create a new
infrastructure management strategy based on emerging technologies and innovative processes.
1. Introduction
Infrastructures are the fundamental facilities and systems required to support societal activity
within a certain region. In particular, the transport infrastructure consisting of mainly roads,
bridges and tunnels, is one of the oldest and most crucial elements for society as it embodies
the physical platform for the transportation of passengers and goods [1]. Since deterioration of
the transport infrastructure poses a serious public safety issue and has a negative impact on
the nation’s economy, to effectively maintain the performance of the transport infrastructure
is of utmost importance. Nevertheless, the accelerated degradation of structures caused by
SynerCrete’18 International Conference on Interdisciplinary Approaches
for Cement-based Materials and Structural Concrete
24-26 October 2018, Funchal, Madeira Island, Portugal
prolonged exposure to harsh environments added to the advanced age of many structures and
the ever-increasing level of demands in terms of traffic loads, volume and vehicle speeds [2],
represents an enormous challenge for the effective management of the transport infrastructure,
which is often translated into an increased frequency of inspections and maintenance
operations. Due to the conservative nature of the construction industry relying on traditional
methods and lagging behind in innovation compared to other sectors, these inspections are
today very inefficient and, together with maintenance operations, constitute a major part of
the recurrent costs of infrastructure, which represent a significant share of the annual budget
in developed countries [3].
New management methodologies must, therefore, be developed based on newly available
technologies and new ways of thinking, to deliver smart systems that could minimize the
number of interventions in the transport infrastructure as well as their extent and duration.
Along these lines, current technological development is taking society to a new era where an
unprecedented type of knowledge will be accessible through the combination of: (i) extensive
wireless sensor networks; (ii) ubiquitous and remote access to cloud-stored data and services;
(iii) constantly increasing computational power and more efficient calculation algorithms; and
(iv) novel tools to visualize digital information. The successful utilization of the
aforementioned technologies will enable a whole new range of possibilities such as a sensor-
driven cloud-based strategy for infrastructure management. The main aim of this paper is to
investigate the feasibility of combining emerging technologies to create an integrated system
that enables an upgrade of the current infrastructure network through a new generation of
smart structures. The study focuses on the aspects that are required to ensure that such a
system could be successfully implemented in practice.
2. Review of emerging technologies
In order to create a system that minimizes the need for structural inspections, some sort of
Structural Health Monitoring (SHM) system must be implemented. A SHM system is a
technology based on the continuous condition assessment of a structure through the analysis
of data acquired on-site by a distributed network of sensors [4]. If properly implemented,
SHM systems could extend the service life of structures by ensuring the early identification of
deterioration and/or damage, thereby allowing relatively minor corrective actions to be taken
before the damage grows to a state where major actions are required [5]. For a SHM to be
effective a streamline of data must undergo four different steps: acquisition; management;
analysis; and visualization. In the following section, a review of the different existing
technologies enabling these four steps is presented.
2.1 Data acquisitionSensor networks
Sensors are critical elements in SHM systems, which must be chosen adequately to serve the
intended purpose under the expected conditions and for a certain time span. Sensors can be
subdivided according to various features, e.g. wired/wireless, embedded/external or
active/passive. Each type of sensor possesses its own advantages and drawbacks, which must
be considered with care. Regarding the measured properties, the most commonly used sensors
in SHM applications include sensors measuring kinematic parameters, such as displacements,
SynerCrete’18 International Conference on Interdisciplinary Approaches
for Cement-based Materials and Structural Concrete
24-26 October 2018, Funchal, Madeira Island, Portugal
strains and accelerations, dynamic quantities, such as vibrations and forces and environmental
factors, such as temperature or relative humidity (RH), some of which are commercially
available for concrete applications, see e.g. [6]–[8]. Conventional sensors often present
difficulties to perform stable and reliable readings in the long term. Many sensors can be
easily affected by changes in temperature, humidity, cable length, magnetic or electric fields,
etc, whereas other sensors need to be powered, which requires the use of batteries, thus
limiting the service life of the sensors [9]. Nevertheless, the common problems that are often
encountered with conventional sensors today will most likely be overcome in the future as
new sensing technologies are developed for bridge monitoring and other large structure
applications. Two examples of novel sensor applications, currently under development, which
possess great potential for the long-term monitoring of reinforced concrete structures are
smart cement-based sensor [10] and polymeric optic fibre [11].
2.2 Data managementCloud services
In large infrastructures with distributed sensor networks containing tens or hundreds of
sensors measuring continuously, the amount of generated data can easily surpass the storage
capacity of any modern computer. To manage such data volumes and enhance accessibility to
the content, a series of cloud-based platforms exists, which provide a wide range of services
for users with the only requirement of an internet connection. These platforms are referred to
as PaaS (Platform as a Service) and can be defined as a cloud computing model in which a
third-party provider delivers hardware and software tools (usually those needed for
application development) which can be accessed anywhere via a web browser. One of the
main advantages of a PaaS providers is that they host the hardware and software on its own
infrastructure, thereby freeing users from having to install in-house hardware and software to
develop or run new applications. Some of the best-known PaaS include Microsoft Azure,
Google App Engine or Amazon Web Services.
2.3 Data analysis – Machine learning
Another of the key steps towards the implementation of an effective SHM system is the
analysis of the measured data. Individual data values by themselves are meaningless. They
need to be situated in a context, relativized to other parameters and combined with a certain
set of assumptions to extract relevant information. This information must be placed within a
theoretical background and used in conjunction with a model to obtain knowledge. Lastly,
this knowledge can eventually be turned into expertise, i.e. the required parameters for
decision making, through experience and training. The first two steps are relatively easy to
automate but for the last one, an experienced and trained operator is still required. This could
change in the near future through the implementation of artificial intelligence, i.e. machine
learning algorithms, which could not just become a decision support tool for engineers but
also unlock the path towards predictive structural assessment.
Machine Learning is currently being used in many existing fields of research to develop
countless applications, some of which are fully operational in everyday situations such as
spam filters or face recognition systems. For structural health monitoring applications, two
main anomaly detection approaches, which may classify as precursors of today’s machine
learning, have been previously used for damage identification: model-driven methods and
SynerCrete’18 International Conference on Interdisciplinary Approaches
for Cement-based Materials and Structural Concrete
24-26 October 2018, Funchal, Madeira Island, Portugal
data-driven methods. The former rely on high-fidelity physical models to detect deviations of
the measured data whereas the latter usually adopt a statistical representation of the system
where data appearing in regions of very low density may indicate deviation from normality
[12]. In civil engineering, the application of machine learning has been also attempted,
particularly for vibration-based damage assessment of steel bridges, see e.g. [13]–[16].
However, owing to the large size and one-of-a-kind nature of the transport infrastructure
elements, the development of effective and generic machine learning algorithms for structural
health monitoring have not yet been developed.
2.4 Data visualization – BIM
Effective data visualization is another crucial aspect for the successful implementation of an
integrated SHM system. The information, whether it is raw measured data or a sophisticated
damage index, needs to be conveyed in a clear, efficient and intuitive way to the operator.
Building Information Modelling (BIM), combining 3D computer-aided design visualization
with integrated data, is a process originally intended to improve the performance of building
projects during their construction phase and service life. Due to the high complexity of the
transport infrastructure elements, BIM stands out as a very promising alternative for its
integration within a SHM system. Today, a variety of BIM software is available, including
both more user-friendly and intuitive commercial packages and more flexible, free open-
source programmes. Perhaps, one of the major technical challenges is to find a suitable
interface that enables the effective integration of real-time measured data with a 3D design
model. A very promising solution to this technical challenge is provided by Autodesk FORGE
[17], a connected cloud platform comprised of web services, and technical resources that
allows for the development of customized and scalable solutions.
Augmented Reality (AR) also possesses a great potential for the visualization of data on-site,
which could represent a giant leap in the efficiency of structural inspections. By visualizing
information regarding the real condition of the structure as an overlay displayed on the actual
structure, inspection operators could easily spot the location of deficient elements and focus
on critical elements, notably reducing the time and extent of the inspection and subsequently
minimizing the cost and disruption to the infrastructure users.
3. A scientific roadmap to sensor-driven cloud-based infrastructure management
Based on the four steps constituting a SHM system and the different reviewed technologies, a
roadmap towards sensor-driven cloud-based infrastructure management has been drafted. This
roadmap, referred to as SensIT and presented in Fig. 1 as an infographic, identifies the critical
areas where further research is required and how these areas are interrelated.
Sensor technology: the first obstacle to overcome in the creation of an integrated SHM
system is the deployment of a sensor network that measures different kinematic and
physicochemical parameters to form the basis for remaining steps. Ideally, robust and
stable sensors which are not affected by external stimuli and can surpass the service life of
the parent structure should be developed. This data should be then combined with
environmental information as well as previous damage report in the case of existing
structures to offer a holistic view of the structural condition.
SynerCrete’18 International Conference on Interdisciplinary Approaches
for Cement-based Materials and Structural Concrete
24-26 October 2018, Funchal, Madeira Island, Portugal
Cloud-based services: the integration of existing cloud platforms is a key aspect to ensure
the efficiency, flexibility and scalability of the system. Cloud platforms play multiple roles
as: a means to obtain the necessary storage; a platform to carry out remote calculations;
and a tool to manage, share and access data from virtually any place in the world.
Machine Learning: with the steady gain in computational power, these types of algorithms
have shown their immense potential for pattern recognition and anomaly detection in
multiple areas. Novel algorithms need to be developed focused specifically in the needs of
the construction industry and particularly for structural health monitoring of concrete
structures. Moreover, these algorithms might benefit from reciprocal data retrofitting with
detailed finite element analyses.
Real-time BIM: a digital twin of the physical structure created through BIM can become a
very suitable channel to convey the information related to the structure condition of the
transport infrastructure. Interactive BIM applications with an intuitive, user-friendly and
cross-platform interface should be developed to offer an effective and versatile decision
support tool for the owner/manager/operator of the structure.
Figure 1: Schematic representation of the SensIT monitoring system.
It is worthwhile mentioning that the approach discussed in this paper presents two main
limitations that hinder its full applicability as of today. The first limitation is the absence of
sensors specifically developed for concrete applications, which can provide accurate, stable
and reliable measurements during the entire service life of the infrastructure. The second is the
SynerCrete’18 International Conference on Interdisciplinary Approaches
for Cement-based Materials and Structural Concrete
24-26 October 2018, Funchal, Madeira Island, Portugal
lack of advanced numerical models that can describe the various multi-physics phenomena
involved in the deterioration mechanisms of concrete, including corrosion of reinforcement,
alkali-silica reaction, etc. to support the development of machine learning algorithms.
References
[1] J. Fulmer, “What in the world is infrastructure?,” PEI Infrastruct. Invest., vol.
July/Augus, pp. 30–32, 2009.
[2] J. Leander, A. Andersson, and R. Karoumi, “Monitoring and enhanced fatigue
evaluation of a steel railway bridge,” Eng. Struct., vol. 32, no. 3, pp. 854–863, 2010.
[3] OECD - Organization for Economic Co-operation and Development, “Infrastructure
investment (indicator) - [Accessed on 18 October 2017].” 2017.
[4] J. M. W. Brownjohn, “Structural health monitoring of civil infrastructure,” Philos.
Trans. R. Soc. A Math. Phys. Eng. Sci., vol. 365, no. 1851, pp. 589–622, Feb. 2007.
[5] B. M. Phares, T. J. Wipf, L. F. Greimann, and Y.-S. Lee, “Health Monitoring of Bridge
Structures and Components Using Smart Structure Technology - Vol. 1,” Center of
Transportation Research and Education, Iowa State University, 2005.
[6] Giatec, “SmartRock2,” 2017. [Online]. Available:
https://www.giatecscientific.com/concrete-sensors/smartrock2/.
[7] Electrotech, “Themisto,” 2017. [Online]. Available:
http://www.electrotech.se/Produkter/BetongsensorThemisto/tabid/1768/Default.aspx.
[8] H. Arup, O. Klinghoffer, and J. Mietz, “Long term performance of MnO2-reference
electrodes in concrete,” in Corrosion97, 1997, no. 243, p. Paper 243.
[9] H. Ceylan, K. Gopalakrishnan, P. Taylor, P. Shrotriya, S. Kim, M. Prokudin, S. Wang,
A. F. Buss, and J. Zhang, “A Feasibility Study on Embedded Micro-Electromechanical
Sensors and Systems (MEMS) for Monitoring Highway Structures,” 2011.
[10] F. Azhari and N. Banthia, “Cement-based sensors with carbon fibers and carbon
nanotubes for piezoresistive sensing,” Cem. Concr. Compos., vol. 34, no. 7, pp. 866–
873, Aug. 2012.
[11] Shute, “Optical Fibre Sensor System,” 2017. [Online]. Available:
http://shute.dk/?page_id=107.
[12] C. R. Farrar and K. Worden, Structural Health Monitoring - A Machine Learning
Perspective. 2013.
[13] H. Hao and Y. Xia, “Vibration-based Damage Detection of Structures by Genetic
Algorithm,” J. Comput. Civ. Eng., vol. 16, no. 3, pp. 222–229, Jul. 2002.
[14] K. Worden and G. Manson, “The application of machine learning to structural health
monitoring,” Philos. Trans. R. Soc. A Math. Phys. Eng. Sci., vol. 365, no. 1851, pp.
515–537, Feb. 2007.
[15] S. Soyoz and M. Q. Feng, “Long-Term Monitoring and Identification of Bridge
Structural Parameters,” Comput. Civ. Infrastruct. Eng., vol. 24, no. 2, pp. 82–92, Feb.
2009.
[16] A. C. Neves, I. González, J. Leander, and R. Karoumi, “Structural health monitoring of
bridges: a model-free ANN-based approach to damage detection,” J. Civ. Struct. Heal.
Monit., vol. 7, no. 5, pp. 689–702, Nov. 2017.
[17] Autodesk, “Autodesk Forge,” 2017. [Online]. Available: https://forge.autodesk.com/.
Conference Paper
p>In recent years, significant worldwide research has been conducted regarding the performance assessment of bridges and the concept of performance indicator has been introduced However, there are still significant discrepancies in how these indicators are obtained and used. Simultaneously, it is desirable to achieve processes and methods that are direct, i.e. that measured values are directly compared with projected values over time. This project concerns methods for verification of technical performance requirements. The feasibility study brought together interdisciplinary researchers, consultants, and entrepreneurs to gather knowledge, anchor the research agenda, and implement performance requirements. The project concludes that there is a need for a “Holistic multi-parameter verification/validation system” that relies on the knowledge gained in structural health monitoring research.</p
Article
Full-text available
As civil engineering structures are growing in dimension and longevity, there is an associated increase in concern regarding the maintenance of such structures. Bridges, in particular, are critical links in today’s transportation networks and hence fundamental for the development of society. In this context, the demand for novel damage detection techniques and reliable structural health monitoring systems is currently high. This paper presents a model-free damage detection approach based on machine learning techniques. The method is applied to data on the structural condition of a fictitious railway bridge gathered in a numerical experiment using a three-dimensional finite element model. Data are collected from the dynamic response of the structure, which is simulated in the course of the passage of a train, considering the bridge in healthy and two different damaged scenarios. In the first stage of the proposed method, artificial neural networks are trained with an unsupervised learning approach with input data composed of accelerations gathered on the healthy bridge. Based on the acceleration values at previous instants in time, the networks are able to predict future accelerations. In the second stage, the prediction errors of each network are statistically characterized by a Gaussian process that supports the choice of a damage detection threshold. Subsequent to this, by comparing damage indices with said threshold, it is possible to discriminate between different structural conditions, namely between healthy and damaged. From here and for each damage case scenario, receiver operating characteristic curves that illustrate the trade-off between true and false positives can be obtained. Lastly, based on the Bayes’ Theorem, a simplified method for the calculation of the expected total cost of the proposed strategy, as a function of the chosen threshold, is suggested.
Article
Full-text available
Vibration-based methods are being rapidly applied to detect structural damage. The usual approaches incorporate sensitivity analysis and the optimization algorithm to minimize the discrepancies between the measured vibration data and the analytical data. However, conventional optimization methods are gradient based and usually lead to a local minimum only. Genetic algorithms explore the region of the whole solution space and can obtain the global optimum. In this paper, a genetic algorithm with real number encoding is applied to identify the structural damage by minimizing the objective function, which directly compares the changes in the measurements before and after damage. Three different criteria are considered, namely, the frequency changes, the mode shape changes, and a combination of the two. A laboratory tested cantilever beam and a frame are used to demonstrate the proposed technique. Numerical results show that the damaged elements can be detected by genetic algorithm, even when the analytical model is not accurate.
Article
Full-text available
In broad terms, there are two approaches to damage identification. Model-driven methods establish a high-fidelity physical model of the structure, usually by finite element analysis, and then establish a comparison metric between the model and the measured data from the real structure. If the model is for a system or structure in normal (i.e. undamaged) condition, any departures indicate that the structure has deviated from normal condition and damage is inferred. Data-driven approaches also establish a model, but this is usually a statistical representation of the system, e.g. a probability density function of the normal condition. Departures from normality are then signalled by measured data appearing in regions of very low density. The algorithms that have been developed over the years for data-driven approaches are mainly drawn from the discipline of pattern recognition, or more broadly, machine learning. The object of this paper is to illustrate the utility of the data-driven approach to damage identification by means of a number of case studies.
Book
Understanding the rheological properties of fresh concrete, the hydration phenomenon of cement responsible for structuration, the relationship between the characteristics of the porous solid obtained and its mechanical performances or resistance to the aggressive penetration requires a complex knowledge of the physicochemistry of reactive porous materials. The development of simple formulation rules therefore requires the assimilation of this knowledge and a good command of the properties of these materials. The purpose of this book is to provide the mix designer with useful knowledge on granular materials and porous materials, which will enable the innovative design of concrete. Topics covered include the characterization of granular materials, the concepts of porosity and specific surface area, and the transport properties (diffusion and permeation) of concrete. Some of these topics are already covered in other general books dedicated to granular or porous materials. The objective here is to bring them together in one book by adapting them for use by concrete specialists. Applications in the form of exercises are offered at the end of each chapter to enable readers to assimilate the theoretical knowledge and to apply such knowledge to concrete problems encountered in civil engineering.
Article
Electrically conductive cementitious composites carrying carbon fibers and carbon nanotubes were developed and their ability to sense an applied compressive load through a measureable change in resistivity was investigated. Two types of cement-based sensors, one with carbon fibers alone and the other carrying a hybrid of both fibers and nanotubes, were considered. Direct comparisons were also made with traditional strain gauges mounted on the sensor specimens.Sensing experiments indicate that under cyclic loading, the changes in resistivity mimic both the changes in the applied load and the measured material strain with high fidelity for both sensor types. The response, however, is nonlinear and rate dependent. At an arbitrary loading rate, the hybrid sensor, containing a combination carbon fibers and nanotubes, produced the best results with better repeatability.
Article
Vibration of a new concrete bridge was monitored and change in the bridge structural stiffness was identified accordingly over a 5-year period. This three-span 111-m long bridge is instrumented with 13 acceleration sensors at both the superstructure and the columns. The sensor data are transmitted to a server computer wirelessly. Modal parameters of the bridge, that is, the frequencies and the modal shapes were identified by processing 1,707 vibration data sets collected under traffic excitations, based on which the bridge structural parameters, stiffness and mass, and the soil spring values were identified by employing the neural network technique. The identified superstructure stiffness at the beginning of the monitoring was 97% of the stiffness value based on the design drawings. In the identified modal frequencies, a variation from −10% to +10% was observed over the monitoring period. In the identified stiffness values of the bridge superstructure, a variation from −3% to +3% was observed over the monitoring period. Based on the statistical analysis of the collected data for each year, 5% decrease in the first modal frequency and 2% decrease in the superstructure stiffness were observed over the 5-year monitoring period. Probability density functions were obtained for stiffness values each year. Stiffness threshold values for the collapse of the bridge under the operational loading can be determined. Then the number of years can be assessed for which the area under the proposed probability density functions is greater than the threshold value. So the information obtained in this study is valuable for studying aging and long-term performance assessment of similar bridges.
Article
During routine inspections of the Söderström Bridge in central Stockholm, one of Sweden’s most important railway bridges, cracks were found in the web of the main steel beams. The finding initiated theoretical studies which showed that the cracks developed mainly due to poorly designed connections of the cross beams and out-of-plane bending of the web. The studies also showed an alarming result regarding the remaining fatigue life of the stringers and the cross beams. However, no cracks or other damage have been found on these components during the inspections. To explore the differences between the theoretical indications and the inspected reality, an extensive monitoring program has been performed. This article describes the monitoring program and the analysis methods used. Some results regarding the remaining fatigue life based on measured and theoretical values are presented.
Article
Structural health monitoring (SHM) is a term increasingly used in the last decade to describe a range of systems implemented on full-scale civil infrastructures and whose purposes are to assist and inform operators about continued ‘fitness for purpose’ of structures under gradual or sudden changes to their state, to learn about either or both of the load and response mechanisms. Arguably, various forms of SHM have been employed in civil infrastructure for at least half a century, but it is only in the last decade or two that computer-based systems are being designed for the purpose of assisting owners/operators of ageing infrastructure with timely information for their continued safe and economic operation. This paper describes the motivations for and recent history of SHM applications to various forms of civil infrastructure and provides case studies on specific types of structure. It ends with a discussion of the present state-of-the-art and future developments in terms of instrumentation, data acquisition, communication systems and data mining and presentation procedures for diagnosis of infrastructural ‘health’.