website statistics
skip to primary navigationskip to content

Cambridge Centre for Smart Infrastructure and Construction

An Innovation and Knowledge Centre funded by EPSRC and Innovate UK

Studying at Cambridge

 

Data driven decision making

Current 'Data Driven Decision Making' projects

1. Data-centric engineering and bridge monitoring

The following projects also appear under the 'Managing and Operating Infrastructure' theme

2. Data-driven asset management – a framework for linking ISO and BIM standards for whole-life value

3. Automating the visual inspection process for masonry arch bridge condition monitoring

4. Optimal monitoring of a masonry arch skew bridge

The following project also appears in the 'Use and Development of Sensors' theme

5. Development of a remote-controlled boat for underwater surveying  

 

 

1. Data-centric engineering and bridge monitoring

Instrumented infrastructure containing advanced sensor networks can provide the important information required to help transform design, construction, assessment, and maintenance of bridges. However, one of the main challenges facing engineers lies in the development of innovative processing and analysis methods to distil the vast quantities of data collected to useful information. A recent collaboration between engineers at CSIC and statisticians at the Lloyd’s Register Foundation funded Programme on Data Centric Engineering (DCE) at the Alan Turing Institute is focussed on addressing this challenge. DCE synthesises physics-based models, which are updated based on measured data from the actual physical asset, and statistical (data-driven) models to study physical engineering assets. This approach combines physical prior knowledge with empirical data to create a 'Digital Twin'.

To date, CSIC researchers have collected performance data, during construction and operation, from two new railway bridges instrumented with a network of fibre optic sensors. In addition, performance data (dynamic strain and displacement) from an ageing masonry rail viaduct is being continuously collected over an extended period of time. These two data sets provide a rich resource of information for structures at opposite ends of their service life, providing engineers and statisticians with opportunities to transform the current way we think about structural health monitoring. Specifically, several research activities are planned or underway, including:

• Developing advanced statistical models for analysing, visualising and interpreting large and continuously updated monitoring data sets in real-time

• Integrating combined physics-based (e.g. finite element) and statistical models with the aim of detecting, locating, and quantifying structural deterioration and damage while accounting for uncertainties associated with the collected data and models

• Investigating the scalability and portability potential of the proposed methods for implementation on other self-sensing structures.

Project contacts are CSIC Investigator Dr Matthew DeJong, CSIC Research Associate Dr Liam Butler, and CSIC Research Associate Dr Andrea Franza.

 

2. Data-driven asset management – a framework for linking ISO and BIM standards for whole-life value

As the world of BIM L 3, Digital Built Britain (DBB) and UK Digital Economy beckons, there will be a strong focus of data driven construction solutions. This research project, in partnership with Costain, will progress the use of BIM as the cornerstone of information management for asset maintenance and management (BIM levels 3 and 4).  The information required for through-life management and a method to link this to the BIM model to provide a fully integrated asset management platform will be identified for a given asset type.

The aim of the project is to create a model based framework approach to aid in the development of whole-life asset information requirements (AIR) linking the BIM 1192 standards with the ISO 55000 standards (Figure 1).  A tool will be built that can automaticity link AIR to Uniclass 2015 - a unified classification system for the construction industry. Uniclass 15 contains consistent tables classifying items of all scale from a facility such as a railway through to products such as a CCTV camera in a railway station. The asset information model will be validated to the organisation information requirements and objectives.  Project contacts are Dr Ajith Parlikad, Senior Lecturer in Industrial Systems at the Institute for Manufacturing and CSIC Investigator, and PhD Researcher James Heaton.

 

3. Automating the visual inspection process for masonry arch bridge condition monitoring

Currently, the condition of masonry arch bridges is predominantly assessed via manual visual inspection. With the collection of image and laser scan data becoming increasingly fast, there is potential to automate this process. leedsarchresize.jpgThis would help to reduce the subjectivity of the current manual methods as well as reducing cost and improving safety. The automated process being developed seeks to combine image and laser scan data to detect and identify defects in the structure, and to determine their severity. The defects identified will then be used to infer the cause and mechanism of damage in the bridge. In this way, the severity and structural implication of the defects and deformations can be determined to more accurately assess the current condition and capacity of the bridge.  Project contacts are CSIC Investigator Dr Matthew DeJong, and PhD Researcher Daniel Brackenbury

 

4. Optimal monitoring of a masonry arch skew bridge

Many ageing masonry bridges are in need of continuous assessment; this is typically completed through visual inspection, though in some cases crown displacements are monitored. This project seeks to monitor a specific masonry arch skew railway bridge with several techniques, with the specific aim of identifying pros and cons of various monitoring techniques for masonry bridge typologies. Specifically, CSIC, in collaboration with AECOM and Network Rail, will monitor a masonry arch skew bridge that has experienced severe distress. Numerous sensors will be deployed, including fibre-optic sensing and videogrammetry. 

The data will first be used to reveal the in-plane flow of force through the skew arch, which until now has not been well understood.  Second, local strain and displacement measurements will be correlated to global (3D) movements obtained from fibre-optic and videogrammetry monitoring. This will facilitate the decision-making on how to optimally monitor to detect the specific service performance that may be a concern. Interpretation of these results will be complemented by modelling of the bridge, so that the monitoring data can be benchmarked against currently applied modelling techniques.  Project contacts are CSIC Investigator Dr Matthew DeJong, and PhD Researcher Sam Cocking 

  

 

5. Development of a remote-controlled boat for underwater surveying

Although great progress has been made in unmanned aerial vehicles (drones), the technology concerning the underwater or water-surface vehicles have been lacking. This project aims to develop a remote-controlled boat prototype to map underwater elevations. Such a device will be useful in the future for measuring the georeferenced river cross-sectional shapes for flood risk analysis. A twin-hulled boat has been built, with connecting framework and top platform providing room for mounting various instruments. The hulls are hollow, which act as the battery compartment. The heavy battery units placed at the bottom of the boat increase the stability of the boat. Most of the electronic devices are housed in a sealed plastic box on top of the middle platform, including a Raspberry Pi for data logging and a GPS sensor. In order to achieve desired accuracy, single frequency differential GPS operation is used, with an accuracy of a few centimetres. A second GPS sensor sits on a tripod at the base station, which supplies real-time satellite correction to the rover GPS sensor. The sonar is capable of measuring depths of up to 1000 feet. Power supply and extra room are provided to attach an Acoustic Doppler Velocimetry (ADV) to the boat, so that flow velocities can be measured. A Leica single prism precision reflector can be mounted on top of the boat, so that the boat’s position will be double-checked by a total station on the bank. The boat is controlled via a handheld 2.4Ghz remote control, and steering is achieved via the differential thrust. We will further test the boat’s performance, and will use it to measure the bathymetry of the River Cam. Project contact CSIC Investigator Dr Dongfang Liang, and Nathanael West, UROP Student.