Journal of Rock Mechanics and Geotechnical Engineering 7 (2015) 207e212

Contents lists available at ScienceDirect

Journal of Rock Mechanics and

Geotechnical Engineering
journal homepage: www.rockgeotech.org

Full length article

UCIMS: Advances in geotechnical construction and performance

monitoring
Raphael Siebenmann a, *, Hai-Tien Yu b, Robert Bachus a
a
Geosyntec Consultants, 1255 Roberts Boulevard, Kennesaw, GA, USA
b
Specto Technology LLC, 295 Princeton-Hightstown Rd, Suite 11-356, West Windsor, NJ, USA

a r t i c l e i n f o a b s t r a c t

Article history: The Crossrail project currently under construction in Central London has been described as “The Big Dig
Received 15 January 2015 on Steroids”, obviously referencing the Central Artery/Tunnel project in Boston completed in 2007. To
Received in revised form address the multiple demands for timely construction performance monitoring, Crossrail envisioned the
5 March 2015
underground construction information management system (UCIMS) to monitor construction progress
Accepted 6 March 2015
Available online 18 March 2015
and structural health along the entire route, with a network of geotechnical instruments (i.e. slope in-
clinometers, extensometers, piezometers, etc.) and tunnel boring machine (TBM) position information.
The UCIMS is a geospatially referenced relational database that was developed using an open source
Keywords:
Tunneling
geographic information system (GIS) that allowed all stakeholders near immediate feedback of con-
Geotechnical data management struction performance. The purpose of this article is to provide a brief history of geotechnical and
Geographic information system (GIS) structural monitoring software, to describe the structure and operation of the UCIMS, and to demon-
Web-based user interface strate how the functionality afforded by this system provided the requisite feedback to the stakeholders.
Examples will be given regarding how the data management and visualization concepts incorporated
into the UCIMS advanced the geotechnical construction industry.
Ó 2015 Institute of Rock and Soil Mechanics, Chinese Academy of Sciences. Production and hosting by
Elsevier B.V. All rights reserved.

1. Introduction machines (TBMs). The UCIMS was designed to provide real-time

access for all stakeholders to the construction and instrumenta-
An underground construction information management system tion data, and it required an intuitive, map-based user interface,
(UCIMS) was developed to monitor construction data from the new showing progress and data locations relative to the surface features
east-west rail link across London known as the Crossrail. The and other spatial information.
Crossrail project construction works, one of Europe’s biggest con- Numerous web-based geotechnical data management systems
struction projects, with projected costs of £15.9 billion, include exist today to monitor real-time data, but these systems often use
underground development, comprising of shafts, tunnels, and either photos or project drawing computer screen captures as a
portals that extend beneath the already congested London under- base map, instead of digital project map files or live web mapping
ground. Given the urban environment in which the tunneling and layers, such as Microstation, AutoCAD, geographic information
other construction activities are occurring, monitoring of the con- system (GIS) files, or Google Earth imagery data, to put monitoring
struction works and the impact on adjacent underground and data in a geospatial context.
aboveground structures was a key activity for the project man-
agement and engineering teams. The UCIMS monitors tens of
thousands of real-time geotechnical instruments and structural 2. Why monitor?
sensors as well as overall construction progress. In addition to
performance monitoring of a host of construction activities, the The traditional geotechnical engineering rationale for moni-
UCIMS also tracks the progress and performance of tunnel boring toring structural health before, during, and after construction is to
provide a warning system to prevent critical failure of an asset, and
to reveal unknowns and reduce uncertainty inherent in geotech-
* Corresponding author. Tel.: þ1 678 202 9555. nical work. With boring tunnels under some of the most expensive
E-mail address: [email protected] (R. Siebenmann). real estate in the world, however, the geotechnical engineering
Peer review under responsibility of Institute of Rock and Soil Mechanics, Chi- perspective is not the only view to consider. From an overall project
nese Academy of Sciences.
1674-7755 Ó 2015 Institute of Rock and Soil Mechanics, Chinese Academy of Sci-
perspective, performance monitoring is also necessary to inform
ences. Production and hosting by Elsevier B.V. All rights reserved. stakeholders, satisfy regulators, document performance, assess
http://dx.doi.org/10.1016/j.jrmge.2015.03.002 damages, and possibly to reduce litigation.
208 R. Siebenmann et al. / Journal of Rock Mechanics and Geotechnical Engineering 7 (2015) 207e212

From a quality management perspective, monitoring is a

critical part of the quality control and improvement process.
Because of the removal of over 6 million tons (Carhart, 2013) of
earth from London, it was known that some settlement would
occur. While a number of mitigation measures, such as
compensation grouting through grout shafts, have been imple-
mented, monitoring the settlement in the vicinity of critical
infrastructure is still necessary. This monitoring is needed to
compare expected settlement to actual settlement and to eval-
uate the need for additional mitigation measures.

3. A brief history of geotechnical and structural monitoring

software

As instrumentation programs have grown larger and more

complex, with ever increasing amounts of data being generated,
geotechnical instrumentation monitoring data management and
Fig. 2. Second generation system from the early 2000s.
reporting systems have evolved over time to meet changing project
requirements.
The first generation of software can be dated back to the early
1990s (Fig. 1) and in some cases even earlier. At this time, there Examples include: ARGUS (Yu, 2011), Atlas (Slope Indicator,
were typically “small scale” projects with tens of sensors. DOS 2015c), ISiteCentral (Nyren et al., 2011), Geoscope, and Vista
based, custom-built graphical representations were developed for Dataview (Thorarinsson, 2011). Example projects include Green
each project. Some examples include Arley Tunnel, London Bridge Park, Blackfriars, #7 Subway Extension, 2nd Avenue Subway, East
Bassett convergence system (BCS) (Bassett et al., 1999), Jubilee Side Access, and World Trade Center Station.
Line Extension Red Cross Way, Heathrow Expressway, Tamar
Bridge, Brooklyn Pepsi Bottling Plant, New York, and Gil-Hodge’s 4. Project requirements
Bridge.
The second generation of software was prevalent through the In the study by Dunnicliff (1999), he provided a broad overview
early 2000s (Fig. 2). Software tended to be used for “medium” size of the geotechnical instrumentation planning and implementation
projects with hundreds of sensors. Microsoft Windows, Access process. He outlined 21 steps to a successful program, of which Step
database based, graphical interfaces were developed with user 19 is to plan data collection, processing, presentation, interpreta-
configurable alarms notification. Example software packages from tion, reporting, and implementation. In order to address this step in
that time include MultiMon (Slope Indicator, 2015a), BCSWin the process, Crossrail developed a detailed technical specification
(Slope Indicator, 2015b) and I-Site. Example projects include: the for the successful management and reporting of the data being
New Parliament Building, Jubilee Line Extension, Harrods, and the collected using UCIMS.
Channel Tunnel Railway Link. UCIMS was envisioned as a common repository for data from all
The third generation of geotechnical and structural monitoring stakeholders, including numerous contractors and sub-contractors.
software is being used in the present day of “large” scale infra- The technical specification called for an instrumentation data
structure projects, with thousands of sensors (Fig. 3). Today’s management system, with an intuitive map-based user interface, to
software tends to be web-based, without the need to install any store and visualize site maps, geotechnical data, instrumentation
software on the user’s computer. These systems were developed records, and construction records from TBMs and spray concrete
using back-end SQL databases and provide significant functionality, lining (SCL) systems. The amount of data to be managed included
including alarm messages, automated PDF reports, and more. They over 35,000 instruments and 79,000 sensors monitoring the earth
are often purchased as software services rather than stand-alone on a minute-by-minute or second-by-second basis (Carhart, 2013).
software. Given the large number of sensors, the system needed to include

Fig. 1. First generation system from the early 1990s. Fig. 3. Third generation system from the present day.
 R. Siebenmann et al. / Journal of Rock Mechanics and Geotechnical Engineering 7 (2015) 207e212 209

search functions to locate and view data, navigate interactively (pan geographic data rendering engine that makes GIS data accessible as
and zoom), identify and show features by selecting station number, web map service (WMS) and web feature service (WFS) layers.
coordinates, and names, and view in both plan and cross-section. In OpenLayers is an open source JavaScript library for displaying map
addition to instrumentation data, the system needed to manage data, such as WMS and WFS layers, in web browsers. UMN Map-
and display stakeholder information such as historical building Server was used as a “back-end” server to process the GIS data and
information, utilities, and other assets. OpenLayers was used as a “front-end” client interface to combine
Ultimately, UCIMS was required to function as a reporting tool and display data to the user.
for stakeholders, such as contractors, local authorities, utilities UMN MapServer was used to convert project data, including
owners, and other interested parties. Each stakeholder was given CAD drawings of both aboveground and underground construc-
access to data from their own areas to monitor performance before, tion works, asset information, and historical building information,
during, and after construction in real-time. into web-accessible data layers. Several other GIS servers were
considered, including ArcGIS Server, GeoServer, Mapnik, and
MapGuide. UMN MapServer was chosen primarily because it has a
5. UCIMS architecture powerful and efficient mapping engine that fits well with the
existing ARGUS stack. It was also chosen because it has strong
UCIMS was developed using itmsoil’s ARGUS monitoring soft- track record of use since the mid-1990s, has an active developer
ware codebase, which was extended to include a GIS interface. A community, and supports a multitude of raster and vector data
system diagram is provided in Fig. 4. The ARGUS codebase consists formats.
of a bundle of open source software that includes Linux, Apache, OpenLayers was used to dynamically display instrumentation
MySQL, and PHP, commonly known as a LAMP stack. In this stack, data, TBM position, and other overlays on top of a variety of base
Linux is the operating system, Apache is the web server, MySQL is map layers. OpenLayers provided the framework for allowing basic
the database management platform, and PHP is the server-side user interactions such as panning, zooming, and turning GIS data
scripting language used to render the web pages. layers on and off. It also allowed for a number of advanced features,
ARGUS provided a powerful web-based graphical user interface such as symbol clustering, to be implemented. OpenLayers was
(GUI) that allowed for the configuration, management, and display selected again because it fits well within the existing ARGUS stack
of large amounts of instrumentation data from a variety of sources and was not “bundled” with other software libraries. Other Java-
(Fig. 5). However, in ARGUS, the sensor data were displayed only script application program interfaces (APIs) were evaluated,
on static images, meaning that for each project numerous indi- including Google Maps API and ESRI’s ArcGIS Javascript API; how-
vidual “plan views” needed to be configured to display sensors. ever, OpenLayers was ultimately chosen because of its optimal mix
These plan view images are typically either photographs or screen of features, documentation, and limited footprint in the overall
captures from CAD software packages that do not allow for the application.
display of complex geometric objects, such as a line representing
the passing of a TBM. Given the existing capabilities of the ARGUS
codebase, the goal of the UCIMS development team was therefore 6. Key functionality
to select and leverage compatible technologies to meet the project
requirements. UCIMS was designed to solve the problems of accessibility, scale,
The GIS interface was developed using a multi-tier client-server and context. A number of innovative tools, technologies, and
web GIS architecture consisting of University of Minnesota (UMN) techniques were used to visualize geotechnical instrumentation
MapServer and OpenLayers. MapServer is an open source data, TBM, and asset information simultaneously in real-time.

Fig. 4. UCIMS diagram.

210 R. Siebenmann et al. / Journal of Rock Mechanics and Geotechnical Engineering 7 (2015) 207e212

Fig. 5. Example ARGUS plan view.

6.1. Accessibility within the geometric envelope of the map. When the user zooms
in, the cluster of points is split into smaller clusters, until the user
Desktop GIS tools have long been able to perform the zooms into a scale where there is enough distance to distinguish
necessary data management and analysis required for a project between individual sensors.
of this scale. However, given the distributed nature of the proj- A robust set of filtering tools were developed to allow users to
ect, with both internal and external project stakeholders in visualize a subset of sensors for each project area. First and fore-
multiple locations, the use of desktop software was not feasible. most, UCIMS allows for the filtering of sensors by alarm status
In addition, while powerful, desktop GIS software requires sig- (high, medium, low, no data, etc.). This filter allows users to quickly
nificant training in order to be used effectively and efficiently. spot potential problem areas and evaluate further. Sensor infor-
The users of UCIMS were not expected to have any formal GIS mation can also be filtered by sensor type (tiltmeter, crackmeter,
training. piezometer, inclinometer, etc.), sensor group (Area A, Area B, etc.),
UCIMS required an intuitive user interface, like Google Maps, and by name. A spatial filter was developed to allow users to select
which most users would need no training to use. It was there- sensors within a certain radius of a point on the map selected by the
fore developed as a “thin” web-based client, with simple, user. Users also are able to combine, save, and load filters for future
streamlined features designed to be responsive for the majority use.
of users. Using a “less is more” approach, the number of base
map layers available was limited to an overall key map, a project 6.3. Context
base map, above ground layers, below ground layers, and Lon-
don underground layers. Optimizing the number of map layers During a major construction project, it is not uncommon for a
reduced the number of clicks needed to display relevant infor- contractor to be closely monitoring a local instrumentation
mation and improved the overall website performance and user network without being aware of nearby work, other instrumenta-
experience. tion, or factors that may affect their local instruments. By
combining instrumentation data into a central repository, UCIMS
6.2. Scale creates a comprehensive picture of existing and historical
conditions.
The issue of scale was addressed using a combination of clus- With the addition of TBM and SCL position information, stake-
tering and filtering. Given the thousands of sensors that needed to holders are able to more quickly discern whether a recent alarm is
be displayed for each project site, without a combination of clus- due to active construction (Fig. 7). UCIMS also incorporates building
tered symbol and sensor filters, users would be easily overwhelmed information and other GIS data layers, such as station numbering,
by the number of sensors on the map and would be unable to and provides access to relevant information that allows stake-
identify meaningful spatial patterns. holders to make informed decisions based on the context of the
A clustering algorithm was applied to group multiple sensors, instrumentation alert and its surroundings.
which means that when sensors are within a specified distance of
each other on the map, they are grouped as a single symbol (Fig. 6). 7. Conclusions
The color of the symbol denotes the highest alert level in the
cluster. The clustered point can then be clicked to inspect the in- UCIMS was developed to provide an instrumentation and con-
dividual sensors within that group. The overall effect is an un- struction monitoring system of unprecedented functionality that is
cluttered map that efficiently displays all the sensors that are accessible through a user-friendly web-based map-driven
 R. Siebenmann et al. / Journal of Rock Mechanics and Geotechnical Engineering 7 (2015) 207e212 211

Fig. 6. Point clustering allows for an uncluttered display of thousands of sensors.

Fig. 7. Current TBM position is displayed in red alongside current instrumentation readings.

interface. Using an open source multi-tier client-server web GIS need to react to the output information and make time-critical
architecture, consisting of UMN MapServer and OpenLayers, an decisions regarding construction activities.
existing geotechnical and monitoring software package was
extended to address the challenges of accessibility, scale, and Conflict of interest
context.
The UCIMS is capable of displaying the output records of thou- The authors wish to confirm that there are no known conflicts of
sands of sensors at a time and overlaying these sensors over base interest associated with this publication and there has been no
maps, including CAD drawings, geo-referenced aerial photographs, significant financial support for this work that could have influ-
and raster maps, as well as photographs. Among many features, enced its outcome.
users are able to pan, zoom, search and filter sensors, store views,
measure areas, and click on sensors to produce time trend plots of
References
data. These time trends, coupled with independent and automated
alarm levels, provide timely feedback and protection to stake- Bassett RH, Kimmance JP, Rasmussen C. An automated electrolevel deformation
holders. The functionality incorporated into the UCIMS not only monitoring system for tunnels. Proceedings of the ICE e Geotechnical Engi-
facilitates information presentation, but, importantly, also provides neering 1999;137(3):117e25.
Carhart N. The crossrail major infrastructure project e high-end risk management
timely information to operators, engineers, and managers who at work. Hazards Forum Newsletter December 2013. No 81.
212 R. Siebenmann et al. / Journal of Rock Mechanics and Geotechnical Engineering 7 (2015) 207e212

Dunnicliff J. Systematic approach to planning monitoring programs using Dr. Hai-Tien Yu has more than 40 years of experience in
geotechnical instrumentation: an update. In: Proceedings of the 5th Interna- the instrumentation industry with consultants, manufac-
tional Symposium on Field Measurements in Geomechanics e FMGM99, turers and contractors. He has played key roles in the
Singapore; December 1999. instrumentation and monitoring components of many
Nyren R, Marr A, Jacobs D. iSiteCentral: web-based data management software. major projects across the world. He is highly respected in
Geotechnical Instrumentation News 2011. the industry and his wealth of technical knowledge has
Slope Indicator. Atlas web-based monitoring. 2015. http://www.slopeindicator.com/ allowed solutions to instrumentation problems to be
atlas/index.php. reached, time after time. Contact information: Tel: þ1 866
Slope Indicator. Multimon & graphx software. 2015. http://www.slopeindicator. 925 7737; E-mail address: hai-tien.yu@spectotechnology.
com/instruments/datalog-multimon.php. com
Slope Indicator. BCSWin. 2015. http://slopeindicator.com/pdf/bcswin.pdf.
Thorarinsson A. Web-based data management software. Geotechnical Instrumen-
tation News 2011.
Yu H. ARGUS: web-based data management software. Geotechnical Instrumenta-
tion News 2011.
Dr. Robert Bachus, P.E., principal engineer based in
Georgia with more than 30 years of experience, focuses on
Mr. Raphael Siebenmann is a civil engineer with Geo- geotechnical engineering, geosynthetic engineering, waste
syntec Consultants in Kennesaw, GA, USA. His experience by-product characterization, and waste containment. He
encompasses site characterization program development provides design, analysis, and forensic investigation on
and management, geotechnical and environmental data projects related to these practice areas.
management and visualization, fate and transport
modeling, and risk assessment.