Implementation of Shaft and Tunnel Excavation Monitoring System in the Deep Tunnel Sewerage System Phase 2 Project

Lai Lynn Woo, Aung Ko Ko Soe, Kyi Khin, Darryl Tan

National Water Agency PUB, DTSS2 Department, Conveyance, Singapore

Angus Maxwell, Elpidio Valdez. Jr

Maxwell GeoSystems, Singapore

ABSTRACT: Singapore’s Deep Tunnel Sewerage System Phase 2 (DTSS2) project comprises 50 km of tunnelling work by 19 Tunnel Boring Machines (TBM) and another 50 km of link sewer pipe jacking work to channel used water to a new centralised water reclamation plant to be constructed at Tuas – Tuas Water Reclamation Plant (TWRP). A number of the tunnel shaft locations also include Air Jumpers (AJ), Odour Control Facilities (OCF) and hydraulic structures. Therefore, construction of DTSS2 will produce significant amount of data every day particularly from well instrumented TBMs and pipe jack machines. Since PUB recognizes the volume of data during construction, the Shaft and Tunnel Excavation Monitoring System (STEMS) is introduced in DTSS2 project to integrate all construction monitoring and tunnelling data into a common system environment. The aim is to deliver better quality and timely data with various analytics and reporting capabilities for decision making and risk management. Therefore, engineering geological data, geotechnical instrumentation data, excavation data, TBM and pipe jacking machine data, construction progress and other metadata across the project are taken into the system for processing and integrated into various user definable formats to facilitate quick review and combined analysis. The web-based and integrated nature of this centralised processed data management system also has real-time TBM monitoring capability and automatic SMS alerting feature. Hazards and risks, which are always linked to ongoing activities and planning, can also be identified ahead of time. This paper focuses on the application and implementation of STEMS and also highlights the importance of collaboration for successful implementation.

1. Introduction

The Deep Tunnel Sewerage System (DTSS), an underground highway for used water management, is a core water infrastructure which provides a cost-effective and sustainable solution to support Singapore’s continued growth and meet its long-term needs for used water collection, treatment, reclamation and disposal. DTSS Phase 1 (DTSS1), comprising the North and Spur Tunnels, the associated link sewers, the Changi WRP and outfall, was completed in 2008. DTSS Phase 2 (DTSS2) comprises the South Tunnel which conveys domestic used water, the Industrial Tunnel for non-domestic used water, associated link sewers and the Tuas WRP.

The DTSS2 conveyance system has 100km of deep tunnels and link sewers of which 50km is constructed by 19 Tunnel Boring Machine (TBM) and 50km by pipe jacking. Deep shaft and manhole excavations are also required to facilitate launching/receiving the boring machines and to construct associated hydraulic structures. The tunnel construction is split into 5 design & build contracts (Contract T-07 to T-11) and link sewer construction by pipe-jacking is split into 3 schedules (Schedule I, II and III) where each schedule is packaged into 3 to 5 build only contracts. Geographically, the tunnel alignment runs largely under major expressway corridors at between 35 to 55 metres below ground and is bounded by transportation infrastructures, educational institutions, healthcare, commercial, industrial and residential buildings. The geological formation prevailing in the project area is Jurong Formation which is notorious for high groundwater inflow during excavation and tunnelling works. Pockets of soft soil deposits from Kallang Formation are also present locally some distance above the tunnel horizon. It is known that excavation and tunnelling process inevitably causes disturbance to surrounding ground and further leads to ground settlement, even induces severe hazards to structures and infrastructures (Yin et al. 2017, Wang et al. 2019, Yin et al. 2020, Zhang et al. 2021). Therefore, all potential risks must be assessed and a comprehensive monitoring regime established to ascertain an acceptable level of security for existing structures in excavation and tunnelling influence zone.

For this important reason, TBMs used in the project are heavily instrumented for efficient operation and better control to minimize the risk of over-excavation which can cause ground collapse and depression on the surface. Likewise, extensive arrays of geotechnical instrumentation are also required to be installed along the tunnelling corridor for monitoring. The substantial quantity of data generated from TBMs and geotechnical instruments require an efficient method for processing and delivering useful engineering data to interpret and assess the level of risks at all times. The most promising and prevalent method to do so is to create a common data environment using web and cloud computing technologies. With reliable and sufficient communication of monitoring data and in-situ construction information on a co-ordinated data platform, it is easy to devise appropriate action plans to avoid major hazards. Therefore, DTSS2, being a major tunnelling project in a densely populated urban area, requires a centralised integrated data management system implemented to deliver better quality and timely data with various analytics and reporting capabilities for decision making and risk management.

2. Background

In urban tunnelling and excavation works, geotechnical instrumentation data and TBM operation parameters are of paramount importance, and close monitoring is vitally important for safe and successful operation. Therefore, centralised database management systems are increasingly utilised in excavation and tunnelling project as they offer the ability to integrate various types of field information and relay this to decision makers in real time. Monitoring data in urban tunnelling projects are generally of two types: real-time data and non-real-time data. Real-time data are data from TBMs, Slurry Treatment Plants (STP), Slurry Transport Systems (STS), Automatic Tunnel Monitoring Systems (ATMS), Real-time Vibration Monitoring Systems, Expressway Monitoring Systems and live CCTV streams. Non-real-time data are manual instrumentation readings, soil investigation bore logs, reports, drawings, records, photographs, etc. Traditionally, these types of data are processed and managed separately on different platforms. Therefore, it is time consuming or operator intensive when combining and integrating this data for analysis.

Nielsen & Koseoglu (2007) pointed out that construction information monitored by manual work was expensive and limited. Moreover, it exhibits the lag effect and thereby cannot assist with the prompt decision making and instruction that is required for tunnelling and excavating work. In fact, there are several monitoring systems used in the industry. However, most of them are only designed to work with specific types of monitoring data and not capable of integrating diverse construction data from various sources. For example, most of the conventional TBM monitoring systems do not incorporate instrumentation and ground condition data or their availability is limited to PDF or non-interactive provision only. When cross-correlation is required, it has to be done manually using spreadsheet or graphic editor software. Depending on the types and volume of data, the process would take from a few hours to even up to a few days in some cases. But with an efficient data management system, it can be done conveniently with a few mouse clicks. Therefore, centralised integrated data management systems are becoming popular in urban tunnelling projects as they also carry many other benefits such as improvement in transparency, quality and timeliness in construction data delivery and accessibility.

Recognizing the recent development of data management systems and its applicability, Singapore’s National Water Agency (PUB) decided to utilize a Shaft and Tunnel Excavation Monitoring System (STEMS) as a centralised data platform in DTSS2 project to collate diverse monitoring and construction data, convert it to user-friendly formats and deliver it with appropriate analysis tools. As excavation and tunnelling performance relates to multiple factors, engineering geological data, geotechnical instrumentation data, excavation data, TBM operation parameters, construction progress and other associated data are taken into the system to facilitate the combined analysis and interpretation. With this system in place, it is able to retrieve the information which would not be readily available in conventional systems or hidden in the diverse field data. Therefore, loss of essential construction information can also be avoided.

3. Project Management

PUB awarded the contract for the provision for the Shaft and Tunnel Excavation Monitoring System (STEMS) to Maxwell-Geomotion Joint Venture (MGJV) in September 2017.The duration of the contract is 89 months in which the first 7 months is given for system development. Before the deployment, acceptance test and system performance checks were carried out with dummy data. The developed system, STEMS, is designed to support 5 tunnel contracts, 12 link sewer contracts, 9 instrumentation contracts and an influent pumping station construction contract with unlimited number of users. Upon the completion of project, all data and software shall be handed over to PUB. The system scope and delivery are managed by PUB and its consultant, Binnie and AECOM Joint Venture.

4. Stems Overview

STEMS is developed as a web-based integrated data management system using GIS concept on MissionOS platform which is designed to work with diverse construction data. This system is neither a desktop application nor a ready-made software. It runs on any web browser and is configured according to DTSS2 contract specifications and project requirements. STEMS is different from conventional instrumentation monitoring systems or TBM monitoring systems as it also functions as a coordinated data management platform which bring together various construction data and transform them into useful engineering information. As such diverse data in their native formats can be combined or integrated because STEMS collect and collate them directly from their sources and process them in a way that everything can be combined and viewed in a single common platform (Figure 1).

Nowadays, availability of megabit-speed internet protocol on large excavation and tunnelling projects in urban areas allows STEMS to interface seamlessly with all key data sources. Real-time automatic data are processed and published in real time and notification/alert messages are triggered automatically when pre-set values are reached or breached. STEMS is built up with many tools and functions that are needed for efficient monitoring. The system has various modules covering the key aspects of construction monitoring such as instrumentation monitoring and data analysis, construction progress monitoring, data visualization, reporting, TBM monitoring, pipe jacking machine monitoring and Vertical Shaft Sinking Machine (VSM) monitoring. Various reports in pre-designed format can also be prepared and generated with readily available information in the system. Other peripheral information or documents can also be uploaded and archived systematically for instant access. A detailed discussion about system architecture and functionalities is given in subsequent sections.

5. System Design And Architecture

STEMS is built using standard web scripting languages on a LAMP (Linux, Apache, MySQL, PhP) architecture for server and client-side processing. The architecture is designed as a component based scalable system allowing a high degree of expandability with flexibility for rapid development without the need for hard coding. It utilizes the Amazon Web Service (AWS) server to provide scalable architecture to manage increasing data volume and processing requirements without interruption. For security, data is archived using rolling daily, weekly, and monthly snapshots and can be restored in less than 2 hours. Moreover, a separate local server is installed, and data is mirrored daily.

As shown in Figure 2, STEMS’s structure has three layers – underground layer, site layer and web layer. The underground layer constitutes the data from shaft excavations and TBMs. The site layer represents the instrumentation data, drawings, reports and records. However, the web layer only has STEMS server or cloud server. In the web layer, all data from underground and site layers are processed, stored and retrieved upon request. Therefore, the web layer is the key component in the structure and subjected to strict exploitation, quality control and operational performance. Since contract specification requires data to be provided at site layer and web layer, the system is setup on several tiers or layers. Each layer can host its own virtual LINUX server which records the latest data from the data source and controls the upload of this data to the web. Data sources which are not directly web addressable require data to be pushed from the site in which case a small local application can manage the process. Services can be mounted on site computers to act as pre-processors of site data sending to FTP servers and then to Linux/Apache web server for processing and updating MySQL databases. The system is totally web-based and runs on Windows, MacOS and Android operating systems. It functions on popular internet browsers such as Explorer, Microsoft Edge, Chrome, Firefox Safari, Brave etc. It is also optimized to use with tablets and smart phones with at least a 3G connection to the internet. No additional software or plugins are required to use STEMS. For security management, three levels of security are implemented to manage the confidentiality and integrity of the data. First level is the system administrator level where user rights are assigned within the application to grant or restrict the data or functionalities to access. At administrator level, users can see all the data including data in error and have the option to quarantine the data for later correction and adjustment. Second level of security management is the operating system and database administration level in which SQL routines or programs are restricted to the users. Third level is the application-level access where industry standard security requirements are applied.

6. Key Features And Functionalities

6.1 STEMS Portal

STEMS portal is made up of a map/drawing window, a data window, graph preview area and menu bar. The main interface in STEMS is the map/drawing window where different types of georeferenced maps/drawings can be configured to display as a background in it. Aerial images, geological maps, street maps, site plans, combined services plans, instrumentation plans, soil profiles, cross sections, TBM drawings, etc., can be brought to display in the map window (Figure 3 & 4). Map view tools include zoom controls, digitizer and distance measurement tools, map selection tabs, legend control, drop-pin button and element selection menu. By default, DTSS2 tunnel and link sewer alignments, shaft and manhole locations, soil investigation boreholes and instrumentation points are displayed with an aerial image background which can be instantly changed over to other maps by toggling. Soil profiles and TBM drawings for every TBM can also be displayed in the map window for quick reference (Figure 4). Visibility of elements in the map window can be controlled by legend setting and they all can be labelled with IDs and/or latest readings. Tunnel alignment shown in the map window is composed of segment ring footprints and these footprints can be painted to visualize the progress or machine operation parameters.

All elements displayed in the map window are interactive. Therefore, key information is outlined in the information pane and can be further explored when an element is selected. For instance, easting/northing/ground level, installation depth, review level, calibration certificate if applicable, installation records and photograph attachments will be shown when an instrument is selected in the map window or in the data window. Similarly, a complete bore log will be displayed in PDF and AGS chart format when a borehole point or an ID is selected. Moreover, pins are also available to mark the locations interactively in the map window. And different shapes of pin can be utilised to indicate shafts, manholes and TBM cutter head intervention locations. Pins can also be coloured to differentiate the construction status such as active, non-active and temporary. In addition, pins also provide a place to park the documents relating to the site or locality and can publish the related information such as daily activity, photographs, records, name and contacts of engineer-in-charge and emergency rescue teams etc. Furthermore, a digitizing tool is also available to interactively draw new elements to identify existing or custom structures to include in 2D sections and 3D models.

6.2 Project dashboard

The project dashboard is a one-stop location which outlines the comprehensive overview of the project. It is designed to provide up-to-date information on construction progress and instruments which have recorded substantial readings. The dashboard tabulates the progress for every shaft/manhole excavation, TBM and pipe jacking drives. The statistics outlined for shaft excavation progress include current excavation level, target excavation level and percentage of excavation completed. For that of TBM drives, statistics shown in the dashboard include current drive status, ring number, head chainage, completed distance, percentage of completion, progress bar and production summary (Figure 5).

TBM daily production chart can also be obtained for any time period. Links are also provided to display TBM locations, soil profile and real-time display of TBM parameters. The TBM progress updates are real time and each individual progress is automatically consolidated to derive the DTSS2 project wide progress as shown in Figure 6. The total tunnelling and pipe jack distances completed, numbers of shafts/manholes being excavated and completed are tabulated and indicated on the map as shown in the figure below. But this option is only available to authorized users.

As mentioned above, the dashboard also enables for the summarization of the list of instruments which breached respective review levels. The list can be filtered to see the breaches which occurred within 24hr, this week or previous week as shown in Figure 7 below. The instrument IDs in the list are cross referenced to instrument locations and will be brought into view when clicked.

6.3 Instrumentation monitoring data

Basically, STEMS takes only raw readings from both real-time and non-real-time instruments. Real- time instrumentation data are automatically pushed forward to STEMS by respective web servers. Other than manual uploading, non-real-time data can also be transferred electronically through web interface or by emailing to a dedicated email server using a provided template file. Incoming raw data are instantly processed according to the respective data models configured in the system. Data models are in fact the calculation templates and every instrument type has their own data model. They are used to translate the output results into various units and formats between the graphing engine and the database. Therefore, graphs can be plotted in various engineering units and allow for integration with other instrument readings or excavation depth (Figure 8b). Other than chart format, readings can also be presented in table and text format (Figure 8c and 8d). During the processing, rate of change is also computed to detect spikes and notify the originator for confirmation to avoid erroneous readings. In addition, STEMS’s graphing tool has various options to display the readings. For quick review, instrumentations points can be labelled with their latest readings in the map window (Figure 8a). Multiple instruments can be plotted together and there is no limit to the number of instruments which can be plotted together in a single graph. Likewise, different types of instruments can be plotted together using multiple axes option for trend comparison. Moreover, it also enables the instrumentation trend plots to be overlaid with the construction activity timeline for correlation (Figure 8b).

Instruments in arrays can also be grouped and enabled to produce 2D section plots to depict the ground movement in relation to excavation progress. Furthermore, it also enables to produce user-defined cross section in which every object within a user defined distance from a cutting line will be plotted. An option is also available for the section to overlay on the soil profile as shown in Figure 9. It will increase the understandability of the ground movement/water drawn down and enable the users to observe the risk or identify the area of concern more easily.

Furthermore, STEMS’s graphing engine also has prediction capabilities. By applying the rate of change computed during the processing, it is able to predict the time in number of days for instruments to reach the review level and highlight it next to the instrument IDs on the chart as shown in the Figure 10.

6.4 Automatic alert notification

When an instrument reading reaches or breaches pre-set threshold limits, the system will automatically trigger SMS and email messages to a dedicated group of users. Meanwhile, it will highlight the breached instruments in distinct colour for easy identification in the map window – yellow for alert level (AL), pink for predetermined level (PDL) and red for work suspension level (WSL) (Figure 11). In addition, all breached events are registered in the system and event blogs created to start interim communication and share views, opinions and justification. The blog has input fields for description, site activity/observation, other instrument readings in the vicinity, actions taken, QPs’ recommendation and remarks. Every single input given in the blog is logged and forwarded to all designated emails. This opens an internal communication platform regarding the breach and then generates an interim assessment report based on the inputs given. The generated reports are archived in the database and no end-users can edit or delete it. This process is automated and the report format was designed after discussion with all relevant parties across the project.

6.5 Data visualization

There are a few ways to realistically visualize the instrumentation readings and borehole data with different abstraction in STEMS. Interpolation is one of the available features and it uses the Inverse Distance Weighted (IDW) method to draw contours of instrument readings except for inclinometers. In addition, rock head contours or depth to specific soil layers can also be drawn (Figure 12). The contours are solid filled but they can be overlaid on the map.

On the other hand, shafts and tunnels can also be viewed in 3D together with instrumentation points and soil investigation boreholes (Figure 13a). In 3D model, construction status of the shafts and tunnels are indicated by colour. Brown colour indicates the constructed portion whereas white colour indicates the parts which have not been completed yet. Similarly, original (before construction) and current water/piezometric levels are indicated by different shades of blue. Inclinometer deflections and ground condition are also portrayed in the model. Therefore, water level/piezometric level draw down, inclinometer deflection, ground settlement/heave can be viewed altogether with excavation outline and progress schematically. Moreover, horizontal and vertical ground movements recorded at various stages of excavation can also be indicated by using the same colour scheme on 2D section plot (Figure 13b), aiding the ease of identification of any ground movement.

Similarly, ‘E’ and ‘F’ instrumentation arrays for bored tunnel monitoring can be grouped and pre- configured to compute the volume loss based on the latest settlement readings (Figure 14a). Calculation method and applied model used in the system were checked and verified by all project parties during the development stage. On the other hand, the system is also able to identify current TBM location on the soil profile with respect to ring number and chainage (Figure 14b). Therefore, the detailed geotechnical condition of current mining stretch can be obtained easily.

6.6 TBM module

The DTSS2 project deploys 19 TBMs, of which 18 are Slurry Pressure Balance (SPB) machines and one is an Earth Pressure Balance (EPB) machine. There are fiber internet connections available to all TBMs. Therefore, STEMS can connect to TBM data servers from all active TBMs through provided IP addresses, usernames and passwords and streamline the data in real time. The TBM data are continuous and transmitted at every 5 to 7 sec. They are processed and stored in a ring file (last erected ring no is assigned) in the STEMS database. A ring file contains clock synchronized machine data from various sensors on the TBM. This data is designed to display in the main and secondary TBM monitoring tabs (Figure 15).

The composite TBM display tab is designed to give a comprehensive understanding of the TBM operation, Slurry Treatment Plant (STP) and Slurry Transport System (STS) in real time. The specific details on operation of supporting units and systems are configured to display in secondary tabs as shown in Figure 15b, 16 and 17a.

Like other TBM monitoring systems, records of parameters can also be plotted against either time, chainage, net stroke or ring number. These parameters are not only limited to view within the TBM module, they can also be combined with instrumentation data or other information in a single graph over the time series. Therefore, the effect or the influence of TBM operation parameters on ground movement or their correlation can be studied and identified. Moreover, rings footprint displayed in the map window can be colour coded based on average value of any operation parameter. As mentioned above, cutter head intervention locations can also be indicated by designated shape of pins (CHI pin) as shown in Figure 17b. Therefore, all relevant information and documents can be described and attached to CHI pins. Cutter head intervention permits, compressed air dive records, face logs, photographs and cutter head intervention reports etc., are attached to CHI pins for easy reference. This TBM module greatly assists the tunneling and supervision teams in the provision of an efficient monitoring platform.

6.7 Pipe jacking and Vertical Shaft Sinking Machine (VSM) monitoring modules

Several micro tunnelling/pipe jacking drives are required for the DTSS2 link sewer construction. The machines with a diameter of 2m and above have onboard data logger and are connected to STEMS for real-time monitoring (Figure 18a). As the number of sensors and instruments on pipe jack machines are much lower than on TBMs, the parameters streamlined to STEMs are not as extensive. Typically, data from the machine comes in at 5 seconds interval and key parameters are displayed in the composite dashboard tabs. All incoming data are stored in the database and can presented in chart format or retrieved in text format as and when needed. VSM is utilized for 5 of the shaft constructions in Contract T-11. The parameters from two main VSM components – cutting unit and lowering unit are displayed in real time. The VSM machine operation parameters are taken into the STEMS server via a data logging PC at site. The VSM monitoring dashboard design is similar to that of the manufacturer’s control screen but slightly modified to add in a bar chart to show the daily excavation progress. A cumulative chart for overall excavation can also be produced (Figure 18b)

6.8 Daily/Shift reporting

A Shift reporting tool is also provided in STEMS in order to standardize the reporting format among tunnelling contracts in DTSS2 project. By using this built-in tool, engineers can prepare a daily or shift report at the end of the day or shift by simply selecting the relevant activities from the list and timing. But for TBM shift reports, shoving, stoppage and ring building timings will be automatically populated because they are extractable from incoming signals. Therefore, it is more accurate than the manual way of doing it in a spread sheet or a word processing software. Supervisor review and approval process is also fully online and routed automatically. Upon completion, PDF versions will be generated and parked in the designated folder with automatic naming. Another benefit is that the system can utilise the captured activities and timing to compute downtime statistics. Therefore, end users can obtain the overview of the production/standstill time and be able to develop maintenance strategies to minimize the downtime. Figure 19 shows the shift reports generated from the system.

6.9 Others

STEMS also has direct connections with CCTVs installed at various locations and TBMs. The links are established by API interface and the recorded footages last for 48hours. Therefore, construction sites and key locations can be monitored in real time without a need to input another password. Moreover, STEMS is also connected to biometric systems from all 5 tunnel contracts. Thus, manpower on site can also be checked and monitored in real time. In addition, various other metadata such as drawings, soil investigation reports, pre-construction survey reports, CHI reports, weekly interpretative reports, monthly instrumentation reports, weekly meeting minutes, excavation permits, GIBR and TBM protocols reports, etc., are hierarchically saved in the form of a tree menu and can be accessed instantly. This modular menu structure allows systematic documentation and instant availability of documents for reference as and when required. The continuous uploading and instant availability of these documents assists in quality assurance process.

7. Discussions

STEMS shows a high versatility in data collection and transparency in processing. The collection and collation of data from different sources and creating a common data environment allow engineers to focus solely on analysis and interpretation rather than processing of data. Real-time access to TBM operation parameters allows managers to monitor tunnelling operation from anywhere and be able to make prompt and correct instruction. The system is capable of opening and running all real-time TBM monitoring dashboard tabs at any one time without affecting the speed and performance. Currently, it is understood that STEMS is the first web-based centralised monitoring system which has the greatest number of real-time TBM and pipe jack machine connections. Table 1 summarizes the type and number of real-time connections currently established in STEMS. Total data size in STEMS to date is 1.02 Terabytes and projected to reach about 2.5 Terabytes by end of the tunnelling works.

STEMS users are of different levels: engineers, supervisors, managers, project officers, directors and it is therefore designed to support various job roles across the project. STEMS assists project management and construction teams in many ways. Project managers and construction managers can obtain accurate and up-to-date construction progress and be able to visualize it in graphical plots, 2D sections and 3D models. Detailed construction activities, status of TBM drives/pipe jacking drives, their daily/weekly and monthly production can be tracked conveniently. CCTV streams and biometric connections are very useful especially for safety teams to remotely monitor safety lapses and manpower audit. The system also helps instrumentation teams and design teams by automation and the saving of man-hours on processing, reporting and data presentation. Canvas reporting is one of the automations which enables the generation of weekly and monthly factual instrumentation reports. Moreover, various interactive analysis tools are also available and able to generate various integrated cross sections/combined graphs for interpretation and reporting. Likewise, the composite TBM dashboard is found to be very useful as it provides a convenient access to key parameters in real time and presented in user-friendly formats. Similarly, report preparation and sign-off procedures are machine driven and thus a need for paper reports is eliminated. Various reports, drawings, records, permits, photographs etc., are sorted systematically and orderly in respective archive folders for documentation purposes. Therefore, any relevant job role can take part in quality control and document auditing. Likewise, the instant access to these documents is very helpful for meetings and discussion. As such, not much time is needed for preparation before meetings because all relevant documents can be opened in STEMS and presented within a few seconds in the meetings. Latest instrumentation readings, real-time TBM parameters, and volume loss which are key components and subjects of discussion in daily meetings can also be displayed instantly. The induced settlement and TBM operation parameters can be correlated and presented straight away. Furthermore, for instant presentation, interactive canvas tool enables for the creation of a content rich summary page by placement of charts and figures which would be updated with latest readings when reproduced at any later stage. This tool provides significant time savings as well as eliminates a need to export and exchange data. In general, the system interface is user-friendly and even a user who is not familiar with the system can operate it. Furthermore, the alert blogging feature and reporting setup help relevant parties to make interim communication over alerted events and enable for the area of concern as well as potential risks to be highlighted. It has proven an efficient tool to manage alerted events. The protocol and procedures are well supported and assist to close out the report within 24 hours. Due to its real time capabilities and informative features, a monitoring room has been set up at project headquarters and equipped with required hardware next to a conference room to assist the crisis task force in response to emergencies should they occur. The PUB DTSS2 department and their consultant team manage the scope and delivery of STEMS services and promotes the use of STEMS across the project. Meanwhile, it also makes sure that the system is efficient and useful for users from all levels in the project at all time. The team also strictly administers security and access to control, and to ensure the confidentiality of the data. Performance tests and penetration tests are carried out regularly to check the functionalities and system security. Moreover, users are only allowed to see their part of the project and some part of the adjacent contracts in interface areas. The project wide cross contract access is only given to the project directorate office and project management team. Therefore, only authorized users can access all types of data from every contract including TBMs and Pipe jacking monitoring modules, CCTVs and system wide progress dashboard. It should also be noted that efficient system support is vital to ensure the system works well for the project. Contractors are working round the clock and some of the features in the system such as TBM monitoring, reporting, SMS alerting features and other real-time monitoring features are essential for their routine works. Thus, 24hr IT support is essential to immediately resolve the issues and problems pertaining to these key features. For system performance and improvement, the system developer (MGJV) always shows its readiness for optimization and enhancement to make the system more useful. This positive inspiration leads to the development of many new features. Some of the features outlined in previous sections were not included in the original design. These new features were idealised and developed based on feedbacks given by end users during the course of implementation. For example, the dashboard design is the result of collaborative work between client, consultants, and the contractors, and developed on the basis of project requirements. Moreover, close coordination and collaborated effort are essential to ensure the system is updated with the latest information and working in a way as it was intended. Regular meetings are required and found to be an effective platform, not only for gathering feedbacks from end users but also to follow up and enforce the outstanding tasks to be done by site teams or instrumentation teams. Therefore, representatives of client, consultants and contractors meet every month (forthrightly during development stage) to identify problems and issues and to discuss and agree on target, follow up actions and timeline. Most importantly, the technical support team is also involved in daily and weekly instrumentation/ tunnelling meetings so that issues pertaining to STEMS can be highlighted and get resolved quickly as instrumentation data presentation and TBM parameter reviews are carried out by using STEMS in the meetings. Not limited to the above, PUB management also track and monitor the status and performance of the system through weekly and monthly meetings and provide guidelines, ideas and advice for enhancements and improved efficiency. Attributing to support from all levels, experience to date shows that the system works well for the project in terms of the efficiency of monitoring and data management in DTSS2 project.

8. Conclusions

This paper presented the architecture, functionalities and some key benefits of implementing STEMS in DTSS2 project. The STEMS has proven its effectiveness in managing and delivering substantial quantities of monitoring and construction data from different sources. Data transparency is very much improved with more reliable accuracy and loss of information is avoided or minimized. It provides one source of truth for monitoring and tunnelling data which is important for dispute management and possible legal issues in the future. STEMS has turned out to be more than an instrumentation monitoring system or TBM monitoring system as it provides optimization and integration potentials among various sets of construction data. As a result, it has been found that workload and time associating to manually combined these data are eliminated. Moreover, digitalisation of construction data with centralised storage offers instant access which enables construction managers to make fast and correct decisions and thus positively impact the construction efficiencies. Therefore, the use of STEMS can be attributed to tolerable risk objective and the need for transparency between all parties involved.

9. References

• Nielsen, Y., Koseoglu, O., 2007. Wireless networking in tunnelling projects. Tunnell. Undergr. Space Technol. 22, 252–261.
• Wang, Z., Yao, W., Cai, Y., Xu, B., Fu, Y., Wei, G., 2019. Analysis of ground surface settlement induced by the construction of a large-diameter shallow-buried twin-tunnel in soft ground. Tunnell. Undergr. Space Technol. 83, 520–532.
• Yin, Z.Y., Wang, P., Zhang, F.S., 2020. Effect of particle shape on the progressive failure of shield tunnel face in granular soils by coupled FDM-DEM method. Tunnell. Undergr. Space Technol. 100, 103394.
• Yin, W.N., Burd, H.J., Martin, C.M., 2017. Finite-element modelling for the assessment of tunnel- induced damage to a masonry building. G ́eotechnique 67, 780–794.
• Zhang, P., Chen, R.P., Dai, Tian., Wang, Z.T., Wu, K., 2021. An AIoT-based system for real- time monitoring of tunnel construction. Tunnell. Undergr. Space Technol. 109, 103766.