2011 Page 144

CONTENTS ACKNOWLEDGEMENTS

The Editorial Board, DSTA Horizons Editorial 2

Chairman Technical Editor Risk Assessment and Horizon Scanning 4 Teo Chin Hock Professor Bernard Tan Experimentation Centre Deputy Chief Executive (Strategic Development) Department of Physics, Faculty of Science National University of Singapore System Safety in Guided Weapon and 16 Members Peer Reviewers Armament Applications Pang Chung Khiang Director (DSTA Masterplanning & Systems Teo Chin Hock Architecting)/Director (DSTA College) Deputy Chief Executive (Strategic Development) Cognitive Systems Engineering Approach to Developing 26

Lee Chee Teng Pang Chung Khiang Command and Control Systems Director (Procurement) Director (DSTA Masterplanning & Systems Architecting)/Director (DSTA College) Chia Lip Teck Enabling Army Learning Transformation through 38 Director (C2IT) Lim Chee Hiong Infocomm Technology Director (Building and Infrastructure) Vincent Lim Buan Sui Acting Director (Human Resource) Wee Kok Ling Introduction to Submarine Design Director (Networked Systems) 48 Chew Keng Cheow Deputy Director (Guided Weapons Chau Chee Chiang & Armament) Deputy Director (National Security) – Secure Mobility for the Enterprise 60 (till 2 March 2011) Alex Lee Yeaw Lip Deputy Director (Systems Engineering) Chew Keng Cheow Capability Development Framework for 74 Deputy Director (Guided Weapons Defence Technology Investments Seah Peng Hwee & Armament) Deputy Director (Sensing & Connectivity) Alex Lee Yeaw Lip Koh Tong Chia Deputy Director (Systems Engineering) Comparative Analysis of Radar and Sonar 90 Assistant Director (Building and Infrastructure) Principles Seah Peng Hwee Quek Bee Tin Deputy Director (Sensing & Connectivity) Senior Manager (DSTA College) Planning and Designing Data Centres Leow Aik Siang 100 Pearly Chua Siew Ting Assistant Director (Knowledge & DSTA Systems) Senior Manager (Corporate Communications) Effects of Degree of Saturation on Ground Shock Dr Koh Hock Seng 110 Too Meng Yuen Head Capability Development (Surface Senior Executive (DSTA College) Warships) Designing Tactical Networks – Perspectives from 120 Jolene Chng Wang Chin Ong Chin Ann a Practitioner Executive (Corporate Communications) Chief Information Officer Desiree Tan Ting Ting Pee Eng Yau Executive (Corporate Communications) Using Analytic Hierarchy Process with Operations 134 Lead Analyst (DSTA Masterplanning & Analysis in Project Evaluation Systems Architecting) 2 EDITORIAL

through Infocomm Technology’ explains how social and rich media applications are employed for learning in the SAF. The article also illustrates how Content Distribution Network technology is deployed to overcome the challenge of delivering rich media content over a limited-bandwidth environment.

There are increasing demands within the defence ecosystem to be productive on the go, but achieving Enterprise Mobility for security-conscious organisations like DSTA is challenging. Pang Chung Khiang To achieve ‘Secure Mobility for the Enterprise’ and enable access to business applications on Editor, DSTA Horizons non-camera mobile devices, several key challenges are involved. The article explains how the Director (DSTA College) solution needs to balance user expectations, security requirements and the capability of current technologies. DSTA Horizons has gained recognition as a reputable technical journal with growing interest and Two articles in this issue show how comparative analysis and technology adaptation can encouraging feedback received over the past six years. It has also become an established journal lead to fresh perspectives and new ideas. First, ‘A Comparative Analysis of Radar and Sonar for DSTA engineers to showcase and share the work that they have done. I am proud to present Principles’ explores the similarities between the two extensively used sensor systems so as to you with the seventh issue. exploit synergistic overlaps in their applications and technologies. Second, ‘Effects of Degree of In the seventh issue, 12 articles were selected with the intention to include a wide spectrum Saturation on Ground Shock’ re-examines existing guidelines on ground shock loading. The of the audiences in our defence ecosystem. These topics cover engineering innovations, new article provides an insight into developing technologies for the mitigation of ground shock initiatives and research, as well as introductory topics. effects on underground structures in Singapore, which experiences high temperatures and abundant rainfall throughout the year. Six articles in this issue present leading-edge technological solutions and frameworks which were developed in response to challenges faced by the defence community. Some of these The remaining four articles are introductory papers to various topics. We have endeavoured not projects are still in the developmental stage, while others have been implemented and have to be extensive but to be holistic, by combining our knowledge gained from literature reviews become success stories today. and our experiences working with the Ministry of Defence and the SAF. It is our hope that this knowledge shared will be useful and applicable to others in the defence community. As the Singapore Armed Forces (SAF) transforms into a Third Generation fighting force, evaluating highly interconnected and interdependent systems becomes a necessity. ‘Using Analytic Capability development has been identified as the key objective of defence technology Hierarchy Process (AHP) with Operations Analysis (OA) in Project Evaluation’ illustrates the use investments. The team proposed ‘A Capability Development Framework for Defence Technology of AHP with the OA methodology to meet the rigours of evaluating complex defence Investments’ and illustrated this framework through historical examples of early defence applications. The article also cites the successful adoption of AHP to various national non-defence technological innovation, and contemporary examples from DSTA. related projects. Armament, protection and speed are key considerations for surface warships. Submarines need With increasing complex interactions between various systems, a more encompassing to achieve these requirements and fulfil even more challenging demands for stealth, range and methodology is also required in the field of Guided Weapon and Armament (GWA). adaptability, while controlling their cost and size. ‘Introduction to Submarine Design’ presents Safety can and has to be managed in the implementation of programmes. ‘System four innovative solutions to meet these critical requirements. Safety in GWA Applications’ describes how the principles of System Safety have been implemented successfully to deliver safer systems to the SAF. The network-centric battlefield today has to cope with unreliable connectivity, limited bandwidth and latency issues due to the tactical operating environment. ‘Designing Tactical Networks – Planning for the future in an increasingly unpredictable environment requires more than Perspectives from a Practitioner’ serves as a guide for the design and configuration of tactical traditional forecasting techniques. To better support analysts in anticipating emerging strategic networks – proposing best practices for implementation and highlighting potential challenges issues, a cognitive-based system was developed at ‘The Risk Assessment and Horizon Scanning of adaptation for application. Experimentation Centre’. The article traces the development of the system and also describes the challenges and the innovative solutions undertaken. Data centres house the most valuable assets of organisations and they require large investments. Organisations have to draw from industry best practices and customise the data centre strategy The pitfalls associated with relying on technology to solve problems in human performance have according to their own business requirements and constraints. ‘Planning and Designing Data sparked interest in using a multi-disciplinary human-centred approach to design and evaluate Centres’ gives an overview of some of the design considerations and implementation approaches complex socio-technical systems. The team demonstrated the efficacy of ‘A Cognitive Systems for data centres. 3 Engineering Approach to Developing Command and Control Systems’ through a cognitive system prototype for human controllers working in the airport traffic control tower. This issue is a testament to the diverse capabilities and expertise that reside within the defence community. I would like to express my sincere appreciation to all the authors for coming forward To keep Generation Y soldiers who are technologically savvy engaged, a paradigm shift is with their contributions. I am confident that with such dedication, DSTA Horizons will continue required in the way SAF teaches and trainees learn. ‘Enabling Army Learning Transformation to be a valuable channel for knowledge sharing. ABSTRACT Risk Assessment Developing a system to aid analysts in anticipating emerging strategic issues is a challenging engineering feat. The and Horizon Scanning Risk Assessment and Horizon Scanning (RAHS) Experimentation Experimentation Centre Centre (REC) has successfully delivered a cognitive-based system to achieve this. This article describes three key challenges for RAHS. First, analyses are based mostly on discovery instead of deduction. Second, there is a lack of suitable commercial solutions. Third, the processes and technology landscape of the analysts are evolving rapidly. To address these challenges, two approaches have been undertaken by the REC to identify and develop RAHS system capabilities to meet the unique analytical needs and to create an REC process to manage the changing operational environment. This article also describes the REC’s vision of how the system can be further enhanced.

Tan Kwan Chong John Kho Yew Hann Joseph Tan Tow Hua Wong Rong Fah Seng Ai Lim Choo Chee Seng Risk Assessment and Horizon Scanning 6 Experimentation Centre

The Government also recognises that organisations such as SPSS Inc. and SAS The modules and tools augment the BACKGROUND effective exploitation of technology is Institute Inc. support business intelligence analyst’s work in collecting and classifying critical in ensuring the success of the and data mining functions. Through information, conducting analysis and We live in a dynamic and unpredictable RAHS programme. Consequently, the REC discussions with organisations doing RAHS reasoning, as well as constructing models world where many elements are was inaugurated on 16 October 2007 to work, the REC also learnt that only a few of and running simulations to anticipate interconnected and interdependent. This spearhead technological exploration, them have dedicated teams of analysts to and discover trends and patterns. This environment presents both opportunities experimentation and capability development anticipate and monitor emerging strategic is done in a collaborative environment and challenges. in support of the RAHS programme. The issues. Among the organisations that have that promotes the diffusion of knowledge REC is currently staffed and managed by such analyst teams, most do not use human across the team of analysts3. Some of Those who can make sense of the chaos will DSTA under the National Security arm of the augmentation systems in their work. The the individual modules and tools are secure an enduring advantage. As Peter Ho, Networked Systems Programme Centre. REC realised that a new cognitive-based explained in relation to the analyst’s former Head Civil Service said at the launch system was needed to exploit new concepts workflow. of the Risk Assessment and Horizon Scanning THE ENGINEERING in RAHS. (RAHS) Experimentation Centre (REC) in Advanced Data Analytics 2007: “We were no longer operating just CHALLENGE Third, there is a constant impetus to update in the simple or complicated spaces, where the RAHS system. The RAHS concept is still There is a wealth of data available today events were more predictable, and the causes Delivering a system that supports RAHS evolving as methodologies are continually and the analyst is challenged to extract the and effects were clearer. Instead, we were analysis does not simply involve identifying integrated into operational processes, relevant data and make sense of it. operating in complex and chaotic domains, and acquiring commercial software. In resulting in new functional requirements. where the first to discern patterns out of chaos fact, complex and challenging engineering RAHS also uses a broad range of technology A typical analyst using the RAHS system starts would have the competitive advantage.” problems are involved. areas such as text analytics and modelling with a database populated with articles of (Ho, 2007) which are developing rapidly. Therefore, interest. These articles have been trawled in First, RAHS analysts operate in a different a robust process is needed to manage the by robots or uploaded manually by fellow Governments around the world have been paradigm that requires systems to support system changes and continually validate the analysts. The analyst can use advanced search challenged with unforeseen situations such the mental processes of discovery rather than system against the analyst‘s operations which tools e.g. concept, pattern and Boolean as the Severe Acute Respiratory Syndrome deduction. Traditional human augmented are constantly changing. search to navigate the database effectively pandemic, the September 11 attacks, and command and control systems are based on and efficiently. For example, concept search the 1997 Asian financial crisis. These are deductive models of reasoning. For instance, Conceptualising, architecting and developing enables the analyst to find related words just some of the more prominent examples in air traffic control or maritime surveillance an effective RAHS system requires engineers or concepts that may be relevant to a of how traditional analytic and forecasting systems, the operator is clear about the who are creative and competent in a range of search query by utilising a knowledge base techniques have failed to identify emerging entities to monitor and their signatures. interdisciplinary subjects including cognitive containing word meanings, syntax, word strategic threats. Deductive systems are suitable in these science and software architecture. Members variations, and relationships between words. mature operating environments where of the REC are highly innovative and open to In a pattern search, query terms are expanded The Government places strong emphasis on there is deep domain knowledge and clear experimenting. to include terms with similar spellings. planning for the future. Scenario planning understanding of the threats. However, was institutionalised in the mid-1990s and the threats identified through RAHS are Article summarisation and timeline tools in has since been embedded in annual strategic evolving rapidly. A RAHS system needs to DEVELOPING SYSTEM the RAHS system provide the analyst with a planning and budget cycles (Ho, 2009). While assist analysts in identifying patterns from CAPABILITIES quick overview of the search results. Thus, scenario planning still serves as a vital tool for seemingly disparate data. This can only be the analyst can easily filter relevant articles forward planning, it is inadequate for dealing done using discovery-based methods which The RAHS system is designed based on a to focus on. The system also has an engine with the inherent dynamism of current machine systems are weak at. cognitive model that is representative of the that can extract and identify entities such complex situations. This is because scenario human thought process. Modules and tools in as organisations, people and locations planning is essentially a linear process that Second, there are no existing RAHS systems the system are categorised under three main referenced across the search material. This extrapolates possible outcomes from existing to refer to. The REC conducted an initial capability areas: helps the analyst to quickly identify the key trends (Ho, 2008). The RAHS programme market survey but did not find any entities associated with a particular domain was initiated in 2004 to build up capability suitable commercial solution. Among the • Advanced data analytics without having to plough through the within the Government to better anticipate systems that the REC has looked at, some • Modelling content manually. Network analysis tools are emerging strategic issues, also referred to as support analytics but they are optimised • Perspective sharing also available to provide the analyst with a 7 ‘black swans’1 and ‘wild cards’2. This capability for specific areas. Systems from Palantir complements the existing scenario planning Technologies, for instance, target finance process. and intelligence clients while systems from Risk Assessment and Horizon Scanning Experimentation Centre 8

Figure 2. Screenshots of RAHS modelling and perspective sharing tools Figure 1. Screenshots of RAHS advanced data analytics tools

visual representation of how these entities be investigated and define relationships The RAHS system is designed to promote phases need not occur sequentially. In fact, correlate with one another (see Figure 1). between the parameters on the basis real-time collaboration across agencies and they often occur concurrently. of internal consistency” (Ritchey, 2009). analysts. Documents, folders and models are Modelling Computational tools allow the analyst to updated immediately to ensure each analyst Explore collapse numerous combinations of possible using the system has the latest information. After reviewing the relevant literature, the scenarios into a handful of combinations that The REC keeps abreast of new technologies analyst will externalise his mental models. are consistent internally. THE TECHNOLOGY that have been developed and applied The Systems Map module allows the analyst INNOVATION globally. However, it is difficult to be at to build network diagrams of variables which Finally, the Ranking Builder module allows the forefront of every field. Hence, the REC are represented as interconnected nodes, the analyst to monitor scenarios or issues FRAMEWORK identifies and focuses on technologies that and to model the causal dependencies among that have been identified. The orchestrator are likely to have high payoffs. This was them. The module also helps the analyst to tool automatically filters articles based The work of the REC goes beyond the difficult initially because the analysts’ identify endogenous reinforcing and on keyword terms into various nodes of delivery of the RAHS system. To develop operations were not clear. Over time, balancing loops within a system, and the the ranking model. The analyst can also and deliver effective human augmentation through the use of DSTA Enterprise factors that have the greatest impact on the define thresholds to monitor the nodes. systems for RAHS analysts, the REC has to Architecture modelling techniques, the REC dynamics in the system (Senge, 1994). The This automation allows the analyst to track continually harness and exploit available was able to elicit and externalise critical RAHS system automates the identification multiple scenarios or issues across different technologies. Thus, the REC created its own analyst workflows. This enabled a better 3 of these loops and the calculation of their domains. unique E Technology Innovation Framework focus on technology domains which have relative magnitudes. It is important for (see Figure 3). been narrowed down from seven to three: analysts to build and share their models Perspective Sharing 3 to enable better collaboration with their The E Technology Innovation Framework • Data analytics colleagues. The Perspective-based Pattern Detection is an iterative process consisting of three • Text analytics module allows analysts to tap the insights phases: Explore, Engage and Enhance. The • Computational modelling Once the dynamics existing in the area of of external stakeholders. This module allows three phases complement and reinforce one concern have been understood, the analyst stakeholders to provide their perspectives another. For instance, technologies identified The REC also charted technology roadmaps can start to formulate future scenarios. The on specific articles of interest. A powerful in the Explore phase drive the formulation for each domain to drive efforts in Scenario Option Space module supports the visualisation tool then consolidates these of experiments, which would involve user technology scanning. This enables the REC 9 application of morphological analysis. In perspectives to assist the analyst in detecting engagement. Requirements identified during to be proactive in recommending new ideas this method, analysts develop a “discrete weak signals and trends (see Figure 2). the Engage phase serve to build the next- when engaging users. These roadmaps are parameter space of future scenarios to generation RAHS system and also provide dynamic and they are refined over time. direction for future exploration. The three Risk Assessment and Horizon Scanning Experimentation Centre 10

oil and natural gas supply chains would affect techniques to enhance situational awareness. local industries and the economy. Thus, the A combined experiment was conducted REC and DSO were led to conduct experiments in 2008 to validate the effectiveness of an and investigate if advanced computational integrated RAHS and Impactorium4 system in modelling techniques could improve analysts’ a tactical setting (see Figure 6). A joint paper understanding of strategic issues. based on this experiment was presented at the Fusion 2010 conference (Svenson et al., An illustration of a highly successful 2010). The REC is working with FOI to further international collaboration that the REC has enhance and refine the use and application maintained is that with the Swedish Defence of these modelling techniques for strategic Research Agency, FOI. Both agencies are analysis. interested in using probabilistic modelling

Figure 3. E3 Technology Innovation Framework

Engage The experiment was completed in 2010 with participants from the Ministry of Information, The REC focuses on operations-technology Communication and the Arts as well as the integration. Through cognitive task analysis, Ministry of National Development. the REC elicits mental models and decision making challenges from the analysts. Sentiment analysis, which applies Based on the identified challenges, the REC computational linguistics to determine an conducts experiments on the relevant author’s attitude towards a topic of interest, technologies that have the potential to is highly relevant in today’s Internet age simplify the job of the analyst. New work where a large amount of information is processes are often developed with the available online. analysts to exploit these latest technologies. Figure 4. Sentiment analysis prototype The REC identified the growing use and The REC conducts experiments using importance of sentiment analysis tools at an two key methods: in-house prototyping early stage through technology scanning, and and development, as well as external quickly proposed a concept of use which was collaborations. Given the small size of the endorsed by the RAHS analysts. The prototype REC, collaborations often involve leveraging developed by the REC is able to highlight and the ongoing research of other organisations. recommend important articles for the analyst Thus, more experiments can be conducted to focus on, saving valuable time. The REC has simultaneously. The risks are also diversified successfully demonstrated the effectiveness when the REC invests resources and of sentiment analysis through the prototype effort in multiple technologies. The REC and it will be integrated into the next version leverages DSTA’s strong ties with local and of the RAHS system. foreign defence research institutions for collaboration. However, if a particular An example of a collaborative effort is the technology capability is required urgently, experiment conducted with DSO National the REC has the ability to conduct Laboratories (DSO) to investigate the prototyping in-house. concept of using interdependency modelling 11 on energy supply chains (see Figure 5). In An example of an in-house experiment this experiment, analysts from the Energy conducted by the REC is the development of Market Authority were interested to Figure 5. Global dependency modelling (energy) project a sentiment analysis prototype (see Figure 4). understand how disruptions to global crude Risk Assessment and Horizon Scanning Experimentation Centre 12

Figure 6. Swedish-Singapore studies of Bayesian modelling techniques for tactical intelligence analysis Enhance Figure 7. RAHS system vision issues regarding the robustness and usability of the system which were addressed in and the use of that picture in directing further Operating in a dynamic field, the RAHS system REFERENCES has to adapt to fast-changing user operations subsequent upgrades. perception and anticipating future events”. and requirements. Enhancing the situational awareness of the analyst is a goal of the RAHS system. Dominguez, C., Vidulich, M., Vogel, E. and CONCLUSION McMillan, G. 1994. Situation Awareness: The RAHS system started out as a web- The REC has identified three main areas Papers and Annotated Bibliography. based system for research and analysis. The current RAHS system provides analysts of experimentation to achieve situational Armstrong Laboratory, Human System As the analysts began using the system, with a suite of tools and modules that are awareness in the domain of RAHS (see Centre, ref. AL/CF-TR-1994-0085. the open-ended nature of RAHS fuelled a more than sufficient to meet their requirement for a flexible layout to facilitate Figure 7). First, an improvement is required requirements. The next generation of the Foo, K.J., Wong, R.F., Ni, D.E., Leong, M.W. data comparison. Thus, the REC decided to in the selection of important articles that the RAHS system by the REC promises additional and Leong W.H. 2007. Developing a Horizon explore portlet technology. However, as the analyst should focus on. Next, key themes analytical and modelling tools as well as Scanning System for Early Warning. Paper modelling requirements grew, it became and hypotheses of the available data streams enhanced usability features. presented at the 12th International Command apparent that a Thick Client was more have to be identified. Finally, there is a need to study computational models that and Control Research and Technology suitable to manage the increased resources Looking ahead, the REC will focus on automatically fuse fragments of information Symposium, Newport, Rhode Island, US, needed for model editing. technologies to further augment the analysts’ into a cognitive knowledge base. This June 19-21. experience by automating processes. The will enable the analyst to concentrate on Anticipating the evolving requirements is REC will also establish systemic approaches verifying the accuracy of these automated Ho, P. 2007. Future Government: Thinking a challenging task. The REC has learnt to to tackle common human analyst failures representations and selecting the areas of about the Future. Speech presented at the use systems architecture to help manage due to cognitive bias and linear thinking. focus. For each of these areas, the REC is Launch of the Risk Assessment and Horizon this uncertainty. Identifying gaps between Finally, the REC will address the challenges of constructing technology and engineering Scanning Experimentation Centre, Singapore, the systems architecture and exploration information overload given the vast quantity roadmaps to identify gaps between current October 16. roadmaps can flag out difficult enhancements and diversity of data available to analysts. at an early stage. and envisioned capabilities. These roadmaps will guide the REC in future experimentation Ho, P. 2008. Thinking about the Future – Cynthia Dominguez et al. (1994) describe Strategic Anticipation and RAHS. National 13 In 2008, the RAHS system was fielded work. situational awareness as the “continuous Security Coordination Secretariat. successfully in an exercise involving extraction of environmental information, participants from various national security integration of this information with previous agencies. The exercise helped to identify knowledge to form a coherent mental picture, Risk Assessment and Horizon Scanning Experimentation Centre 14

Ho, P. 2009. Future Government: Thinking about the Future. Speech at the Fifth ENDNOTES Joseph Tan Tow Hua is Chief Executive Officer of Cap Vista Pte Ltd which Strategic Perspectives Conference, Singapore, 1 invests in innovative technologies that serve the defence and security November 23. Black swans are described in Nassim Nicholas Taleb’s book “The Black Swan” as events with needs of Singapore. Prior to joining Cap Vista, Joseph was Head of the REC. He was a key member of the National Security Engineering Centre and Petersen, J. 1997. Out of The Blue - How to the following three attributes. First, it is an was involved in developing strategic technology roadmaps and providing Anticipate Big Future Surprises. Arlington outlier as it lies outside the realm of regular technological advice to the National Security Coordination Secretariat Institute. First Edition. expectations, because nothing in the past can convincingly point to its possibility. Second, and national security agencies. He held key appointments in the Ministry of Defence (MINDEF) Chief Information Officer Office and was involved Ritchey, T. 2009. Future Studies using it carries an extreme impact. Third, in spite in managing MINDEF’s IT investment portfolio, spearheading the SAP Morphological Analysis. http://www. of its outlier status, human nature makes Enterprise Resource Planning implementation and championing IT swemorph.com/pdf/futures.pdf (accessed 15 us concoct explanations for its occurrence governance initiatives such as portfolio management and enterprise November 2010) after the fact, making it explainable and predictable. architecture. Under the Public Service Commission Scholarship, Joseph obtained a Master of Engineering (Electrical Engineering and Information Senge, P.M. 1994. The Fifth Discipline: The Art 2 Wild cards are described in futurologist Science) degree with Distinction from the Cambridge University, UK, and Practice of the Learning Organization. John Petersen’s book “Out of the Blue – in 1998. Doubleday Business. How to Anticipate Big Future Surprises” as low probability and high impact events that Svenson, P., Forsgren, R., Kylesten, B., would severely impact the human condition. Berggren, P., Wong, R.F., Choo, M.S. and Wong Rong Fah is a Senior Engineer (Networked Systems). He oversaw Kho, J.Y.H. 2010. Swedish-Singapore Studies 3 A more detailed technical discussion the RAHS programme and led a team to implement the RAHS system. He of Bayesian Modelling Techniques for Tactical about these modules and tools can be also played a key role in the establishment and management of the REC. Intelligence Analysis. Paper presented at the found in the paper entitled “Developing a As a recipient of the DSTA Postgraduate Scholarship, Rong Fah is currently 13th International Conference on Information Horizon Scanning System for Early Warning” pursuing a Master of Science (Cognitive and Decision Sciences) degree from Fusion, EICC Edinburgh, UK, 26-29 July. (Foo et al., 2007). Birkbeck University of London, UK. He obtained his Bachelor of Science (Computer and Information Sciences) degree from the National University Taleb, N.N. 2010. The Black Swan. Random 4 Suite of software tools developed by FOI to of Singapore (NUS) in 2000. House Trade paperbacks. Second Edition. create models representing analyst beliefs.

BIOGRAPHY Seng Ai Lim is a Senior Engineer (Networked Systems). She leads experiments in developing prototypes of information extraction and analysis capabilities. Ai Lim also works on trend analysis and sentiment analysis of open source Tan Kwan Chong is an Engineer (Networked Systems). He is in charge of unstructured data. She was involved in exploratory projects in data fusion designing and implementing modelling and simulation experiments at the systems as well as designing and implementing sensemaking technologies Risk Assessment and Horizon Scanning (RAHS) Experimentation Centre (REC). for situational awareness operations. Ai Lim received her Bachelor of A recipient of the DSTA Overseas Scholarship, Kwan Chong graduated with Computing degree from NUS in 2002. a Bachelor of Science (Electrical Engineering) degree from the University of Michigan, US in 2007. He further obtained a Master of Science (Technology and Policy) degree from the Massachusetts Institute of Technology, US in 2009. Choo Chee Seng is a Senior Engineer (Networked Systems) and he is part of the RAHS pioneer team. He is currently overseeing the development of a new system that incorporates advancement of commercial off-the-shelf John Kho Yew Hann is a Senior Engineer (Networked Systems). He oversees analytics tools and evolution in RAHS processes. He obtained his Bachelor of the technical architecture of the REC and ensures that it is adaptable to Science (Computational Science and Physics) degree with Combined Honours exploit new technologies. He is also responsible for scanning and applying and a Master of Technology (Knowledge Engineering) degree from NUS in technologies in the data mining domain as well as fusing data across varied 1998 and 2003 respectively. sources ranging from newspapers to maps and databases. Under the Public 15 Service Commission Overseas Merit Scholarship, John graduated with a Master of Engineering (Electrical and Computer Engineering) degree from Cornell University, US in 2005. ABSTRACT System Safety in System Safety is a methodology used as a tool to deliver safe Guided Weapon and Armament Applications systems to the Singapore Armed Forces. This article follows the development of the System Safety methodology and explains how this methodology has become prevalent in DSTA’s acquisition of guided weapon and armament (GWA) systems. Insights into the methodology are provided using GWA examples.

Tang Chung Hun Yeo Jia Shyang Tan Puay Han System Safety in Guided Weapon and Armament Applications 18

hazards are considered and mitigated reach the destination on time and accepts INTRODUCTION holistically. DELIVERING SAFE the risks of speeding. As a risk mitigation SYSTEMS TO THE measure, he may install tyres with enhanced Safety is embedded in all aspects of Guided In the field of GWA, System Safety is SINGAPORE ARMED road-holding capabilities and check his car’s Weapon and Armament (GWA) work to particularly important due to the potentially FORCES braking effectiveness regularly. The concept deliver cost-effective solutions that are safe, destructive consequences of malfunction of risk quantification and communication of integrated, secure and reliable. Systems or system failures e.g. severe or irreparable residual risks to the end user for acceptance as How does System Safety align with the Engineering, Systems Architecture and System- equipment damage, serious injuries, part of the System Safety process emphasises emphasis on safety by DSTA and the of-Systems (SoS) thinking are examples of permanent disabilities and even fatalities. the importance of enforcing mitigation Singapore Armed Forces (SAF)? How is the engineering methodologies that incorporate The significance of System Safety is well measures on the ground. hazard table formulated carefully by the safety considerations and holistic thinking represented by the GWA community’s motto safety engineer used? Typically, the hazard in their processes. The term ‘System Safety’ – “Mission First, Safety Always”. has become ubiquitous in the field of GWA table is appended to a Safety Assessment System SAFETY IN Report (SAR) for the system in question. The DEVELOPMENT – HAND and is often used to ensure that a systematic, INCREASING COMPLEXITY methodical and holistic framework of safety project team then tables the SAR to a Safety GRENADE THROWING considerations is in place. System Safety AND INTEGRATION OF Board to review the potential hazards and BAY provides the assurance that all safety aspects WEAPON SYSTEMS mitigation measures. have been considered. The re-designing of the Hand Grenade There are many areas that require System Existing weapon standards and guidelines Throwing Bay is a good illustration. The Safety measures and these include used for qualifying munitions are based on project was an early test bed for DSTA’s System SAFETY AND human-machine interface and software a traditional system definition. The scope System Safety framework. Safety measures ITS ORIGINS implementation. Thus, a multi-disciplinary of this definition includes product design, for the throwing bay were deliberated at a approach is required. A System Safety interaction with the natural environment safety review workshop as the members were System Safety began as a preventive Working Group is also needed to assist the (i.e. propellant degradation over time) and familiar with the facility and could contribute measure against mishaps in complex project management team to brainstorm common transportation means. Weapon to the hazard analysis during the workshop. aircraft and missile systems in the aviation all possible hazards. Taking reference from standards and guidebooks exist due to The end product is very similar to the Grenade industry. The System Safety concept calls for the governing standards and guidelines, the long history of the weapon and gun Range in Pulau Tekong today, which is used a risk management strategy based on the safety measures are implemented to manufacturing industry. Experienced by SAF recruits as a rite of passage in National identification and analysis of hazards, as eliminate or mitigate the hazards. In weapon and gun makers as well as original Service. well as the application of remedial controls equipment manufacturers are aware of the considering complex systems such as air using a systems-based approach (Roland and platforms or guided weapon systems, the potential hazards and have refined their There were existing design guidelines for Moriarty, 1990). The US Military Standard number of hazards can sometimes be in designs to mitigate them. Any development, grenade throwing bays but they did not 882D (2000) defines System Safety as the range of hundreds. Thus, rigorous acquisition or systems management meet the training requirements of the SAF. “the application of engineering and consideration and mitigation of all possible engineering based on these existing The earlier version of the throwing bay was management principles, criteria, and hazards are required to make the system standards and guidebooks will provide built as a short U-shaped wall to serve as a techniques to achieve acceptable mishap as safe as possible. Through these thinking engineering solutions within the boundaries shield from grenade fragments that could be risk within the constraints of operational processes, the potential for all hazards of the traditional system definition. projected towards the thrower. If a grenade effectiveness and suitability, time, and cost, is mitigated to “as low as reasonably were to land behind or in the bay, both the throughout all phases of the system life practicable” (ALARP). Existing standards and guidelines can thrower and the safety officer have to take cycle”. provide the basic safety considerations cover on the other side of the wall. They have Residual risks refer to risks which cannot be within a defined system but may not be to run and climb over the wall in a short time The System Safety methodology may seem reduced further as they are often inherent in adequate for the increasing system-to-system frame of four and a half seconds. However, complex but its principles are common in our the activity itself. As part of risk management, integration. Further SoS thinking is required this reaction time is possible only if the daily lives. For example, preventive measures these residual risks and mitigation measures to provide safety analysis outside the set thrower is an experienced soldier. such as reflectors, bollards and speed limits of stand-alone system boundaries. When need to be communicated clearly to the end are implemented at road bends to mitigate users for their acceptance. The case of a car applied at the development or acquisition In reality, our army recruits are only two against the hazard of road accidents. travelling above the speed limit can be taken phase, System Safety is most effective and months into their Basic Military Training Escalators have emergency stop buttons as an example. The driver is aware that he is has a high potential of influencing design regime when they handle the throwing of and built-in fail-safe mechanisms such as exceeding the speed limit and acknowledges 19 – this aids in the incorporation of the a live grenade for the first time. For most braking mechanisms which are automatically the consequences of being caught by a necessary safety features. recruits, this was a challenge despite the engaged during power failures to ensure traffic camera or getting into an accident. numerous drills they had with dummy the safety of commuters. In short, possible He decides that speeding is necessary to

System Safety in Guided Weapon and Armament Applications

20

S/N Hazard Event Potential Potential Initial Recommended Verification Residual Residual Status Description Phase Hazard Causal Risks Control Measure Risks Risks Effect Factors Acceptable

1 Grenade Firing Injury/ Trainee Serious Design: A trial was Medium Safety Closed detonation Death to drops • Construct slope carried out Board within bay personnel armed within the bay to in April 2005 grenade increase to verify in bay probability of effectiveness grenade rolling of grenade into drain bay drain • Construct ditch behind for personnel to jump in to take cover

Procedure: Documented Figure 1. An engineering drawing of the bay design showing the design features the procedure Revise Immediate implemented after the identification of possible hazards. The features protect the Action (IA) drill in the Range • Personnel to Standing personnel if the IA drills are executed correctly. jump into ditch Orders for to take cover compliance • Kicking of grenade is prohibited

2 Grenade Firing Injury/ Trainee Medium Design: Design Medium Safety Closed detonation Death to drops • Construct ditch drawings show Board behind personnel armed behind for safe hazardous bay grenade detonation of fragments behind grenade from an bay • The raised exploding platform prevents grenade in the hazardous ditch do not fragments from have a hitting personnel direct line taking cover in to personnel the bay crouched in the bay

Procedure: Documented Revise IA drill the procedure • Personnel to take in the Range cover in bay Standing when grenade Orders for drops behind bay compliance

Table 1. A System Hazard Analysis table showing possible hazards and recommended control measures Figure 2. The scenario where the grenade drops in the bay (Source: Adapted from the Ministry of Defence) grenades. Should a live grenade drop From the brainstorming session during the accidentally in the throwing bay or behind workshop, a new design was developed it, the recruits may not react fast enough (see Figure 1). With this design, if a grenade to escape from danger. Therefore, the lands in the bay, both the thrower and safety design had to be focused on incorporating officer need to jump into a ditch. If the safety features that offer greater physical grenade lands in the ditch, both of them protection. should stay in the grenade bay, using the raised platform as a shield from the blast. There were no design guidelines for the type Figures 2 and 3 illustrate these possible of grenade throwing bay required by the scenarios. There is no longer a need to run SAF. Hence, a System Safety approach was or climb to safety, reducing the physical adopted at the stage of conceptualising the demands on inexperienced recruits. design (see Table 1). 21

Figure 3. The scenario where the grenade drops behind the throwing bay (Source: Ministry of Defence) System Safety in Guided Weapon and Armament Applications

22

four metres behind them. Figure 4 illustrates offered transparency and gave confidence System SAFETY IN the incident. System SAFETY IN to the users that all conceivable hazards had DEVELOPMENT – LIVES SYSTEMS INTEGRATION – been examined. SAVED Within seconds, 2LT Kok pulled the recruit PRIMUS down and shielded the recruit from the The use of System Safety also led to a more The effectiveness of the re-designed Hand impending blast. The grenade exploded in In addition to eliminating hazards identified sustainable Operations and Support (O&S) Grenade Throwing Bay was demonstrated four and a half seconds. When the fragments at the design stage, System Safety can also phase. This is because potential hazards, on the morning of 8 March 2008. Second- finally settled, both the recruit and 2LT Kok be used to mitigate other hazards during mitigation measures and residual risks Lieutenant (2LT) Kok Khew Fai was the safety emerged from this harrowing experience systems integration e.g. the installation of a associated with the system or equipment officer at one of the four grenade throwing safely. For 2LT Kok’s bravery, he was awarded weapon or gun system on a platform or ship. are documented clearly during the bays at Pulau Tekong Hand Grenade Range. the SAF Medal for Distinguished Act or System Safety is an effective framework in development phase. With this System Safety Upon command, a recruit armed the Pingat Jasa Perwira (Tentera). This incident mitigating interfacing hazards, especially if framework in place, system managers in the grenade, pulled out the safety pin and held proved the effectiveness of the grenade these systems integration projects involve O&S phase are better informed of the critical on to the arming lever. He then released the throwing bay which was designed and multi-disciplinary expertise. design safety, inspection and maintenance arming lever and swung back his right arm conceptualised through the use of System needs. Subsequent modifications can thus to lob the grenade overhead. However, the Safety methodology. The application of System Safety be carried out safely taking into account grenade slipped from his hand and landed methodology in the development of the self- the original safety design considerations propelled howitzer, PRIMUS, has resulted in documented throughout a system’s development. A selfless hero in time of need successful risk mitigation through a better system design. Out of more than 400 potential Posted on Cyberpioneer: 29 Oct 2008, 1600 hours (Time is GMT + 8 hours) hazards identified in the development phase, System SAFETY IN “It all happened so fast, I didn’t really have time to think,” said 2LT Kok. “I just got him flat on the ground, the project team worked with designers, SYSTEM-OF-SYSTEMS lay prone over him and tried to stay very still as we had practised.” safety engineers and users to eliminate all risks in the ‘Serious’ and ‘Medium’ categories. INTEGRATION – FRIGATE This is the first grenade incident to occur since the throwing bays were enhanced with new safety features in 2004, and only the sixth incident in the 41 years since National Service was introduced in Two residual risks categorised as ‘Serious’ AIR DEFENCE SUITE Singapore. remained. One of them involved the timely execution of time-on-target missions. The The integration of the Aster Anti-Missile Deflected shrapnel from the grenade can go as far as 25 metres from the point of impact, but thanks to Missile (AMM) system with the Republic 2LT Kok’s quick thinking and the precautionary high walls of the throwing bay, both men were protected mitigation measure recommended was to from the fragmented shrapnel and escaped with nary a scratch. synchronise the time clocks between the of Singapore Navy (RSN) Formidable-class Command Control Module and the Gun frigates (see Figure 5) is an early example 1. On 2LT Kok’s command to throw the of DSTA’s application of System Safety at armed grenade, the recruit released his grip Operating Module. The other residual risk on the arming lever and swung back his right lies in the recoil mechanism during weapon the SoS level. The traditional approach of arm to lob the grenade overhead. firing – this could be mitigated by having the applying the methodology with focus on 2. The grenade slipped out of the recruit’s the weapon system was no longer sufficient hand and landed four metres behind them, Charger Loader stand in a specific position next to the ditch. with his arms crossed while operating the because the Aster AMM system functions as 3. 2LT Kok immediately pulled the recruit safety interlock. The detailed and systematic part of the larger Anti-Air Warfare defence down into a prone position on the ground suite (or AAW suite). The behaviour of the and covered the recruit with his own body to identification, analysis and mitigation of risks shield him from the blast.

4. The high wall of the throwing bay helped to deflect the blast from the two men.

23

Figure 4. Illustration of the incident at the grenade throwing bay Figure 5. Combat systems on the RSN Formidable-class Frigate (Source: Adapted from Cyberpioneer) System Safety in Guided Weapon and Armament Applications

24

other systems in the suite, such as the Multi- safety assessment, the detailed design and function Radar, Combat Management System the process are invariably scrutinised BIOGRAPHY and Navigation Distribution System have continually to make the methodology work downstream effects on the operation of the better. It also sets the foundation for the Tang Chung Hun is a Senior Engineer (Armament Systems). He is involved in Aster AMM system. As a result of complex development of DSTA’s GWA engineers, the area of range safety such as inspecting and re-licensing existing Singapore interactions among systems, most emergent where System Safety is gradually introduced Armed Forces (SAF) constructed ranges as well as providing technical analysis behaviour were not obvious and had to be and practised. and support for new range projects. Chung Hun works on the acquisition of identified and managed for safe operations. ammunition, conduct of armament trials and review of safety management CONCLUSION policies of the live firing areas in the national weapon effects, as well as on To address the safety concerns and potential projects for range safety and ballistic protection. He also reviewed the safety hazards, a System Safety approach at the suite Since 2000, System Safety has made management policies of the live firing areas in the SAF. In 2007, he was a level was applied from the design stage by significant progress in the process of member of the DSTA Range Safety Team which won the National Workplace a team comprising subject matter experts of acquiring GWA systems and equipment for Safety and Health Innovation Award (Gold Category) for making SAF firing various systems in the suite. In addition to the the SAF. It is a value-added service that DSTA ranges safer. Chung Hun received his Bachelor of Engineering (Mechanical comprehensive safety assessment performed provides to make our systems and equipment Engineering) degree with Honours from the National University of Singapore on the Aster AMM system, a top-level safety safer. DSTA has helped to enhance the (NUS) in 2006. analysis on the functional flow of the suite knowledge and application of System Safety was performed during the development principles and techniques in the engineering phase. Several hazards were identified and community and to a wider audience. The all associated software and hardware System Safety Society (Singapore Chapter) functions, or modes leading to these hazards is an organisation set up for this purpose. Yeo Jia Shyang is a Principal Engineer (Armament Systems). He is involved in were analysed. Frameworks have also been put in place monitoring, inspecting and re-licensing existing SAF constructed ranges. He to encourage System Safety engineers to conducts technical analysis and provides support for new range projects as Following the analysis, several safety related practise System Safety as a professional well as operations and services requirements in the aspects of range safety gaps in the functional flow were discovered. discipline and specialisation. The directives, and weapon effects. Jia Shyang was part of the SAR21 Team, 40mm Air Design changes were implemented to life cycle manuals and dedicated System Bursting System Team and the Comprehensive Maritime Awareness Team eliminate these gaps. In addition, safety- Safety guidelines of the Ministry of Defence which won the Defence Technology Prize Team Award in 2001, and the Team critical functions at the suite level were have been aligned with the guiding principles (Engineering) Awards in 2004 and 2010 respectively. Jia Shyang attained his identified systematically using the Hazard of System Safety. Master of Science (Industrial Systems Engineering) degree from NUS in 2000. and Fault Tree Analysis methodology. These Under the DSTA Postgraduate Scholarship, he further obtained a Master of functions were code-checked, peer-reviewed, In addition to ensuring the robustness of Science (Explosive Ordnance Engineering) degree from the Royal Military closely tracked and verified in greater GWA systems and equipment, System Safety College of Science, Cranfield University, UK in 2002. depth to prevent uncontrolled changes. As will continue to play a critical role in ensuring part of the verification, safety testing was the safety of our servicemen. conducted at the software unit, system and suite levels. Risks that could not be entirely mitigated by design were highlighted as REFERENCES Tan Puay Han is a Senior Engineer (Naval Systems). He oversees the residual risks for acceptance, and appropriate MIL-STD-882D. 2000. Department of Defense acquisition of missile systems for the Republic of Singapore Navy. He recommendations were provided to the users Standard Practice: System Safety. supported IT security research in DSO National Laboratories, and was to further reduce the risk to ALARP. The RSN involved in front-end studies and systems architecting for air defence has successfully conducted two Aster live- Roland, H.E. and Moriarty, B. 1990. System systems. Under the Singapore Government Scholarship (Open) awarded by firings which validated the AAW suite. Safety Engineering and Management. John the Public Service Commission, Puay Han graduated from Cornell University, Wiley & Sons. US with a Bachelor of Science (Electrical and Computer Engineering) degree CHALLENGES IN in 2003, and a Master of Engineering (Electrical Engineering) degree System SAFETY WORK in 2004.

The application of System Safety has its fair share of challenges, mainly in the form of 25 quantifying risks in terms of severity and probability. In the attempt to support the ABSTRACT Cognitive Systems Cognitive Systems Engineering (CSE) is a multi-disciplinary human-centred approach to the analysis, design and evaluation Engineering Approach of complex socio-technical systems comprising people and technology in real-world domains. This article to Developing Command and Control Systems describes an endeavour to investigate the efficacy of the CSE approach to develop Command and Control (C2) systems. A Virtual Assistant and Persistent Sentinels system was developed as a concept demonstrator for the Tower Air Traffic Control operations. The project team adopted the Decision-Centred Design (DCD) methodology in the development of the concept demonstrator. This article describes how the DCD methodology was used to develop design features to overcome three cognitive challenges. Preliminary evaluation was conducted by obtaining expert feedback through a cognitive wall walk, using cognitive indicators to assess whether system features hindered cognitive performance, and observing the work performance of tower controllers in a proof-of-concept trial.

Yeoh Lean Weng Oliver Tan Kok Soon Linus Low Kar Seng Teh Shi-Hua Cognitive Systems Engineering Approach 28 to Developing Command and Control Systems

collaborative skills (Hoffman, Klein and fraught with operational stressors and CSE is a multi-disciplinary human-centred INTRODUCTION Laughery, 2002) in C2 systems development. environmental effects such as ill-structured approach to the analysis, design and problems, dynamic situations, diverse evaluation of complex socio-technical The rapid advancement and adoption of IT This article describes an endeavour to operational goals as well as imperfect and systems comprising people and have radically transformed Command and investigate the efficacy of the CSE approach voluminous information. All these factors are technologies in real-world domains. Control (C2) functions and processes in the to develop C2 systems. The development further compounded by the fast tempo and CSE practitioners combine theories and modern military command post. C2 functions of a C2 system prototype for human tower time-critical nature of C2 operations. techniques from cognitive science, human in the command post are now largely controllers working in the Airport Traffic factors, human-computer interaction design knowledge and network-centric (Yeoh, Mui Control Tower (ATCT) is used to illustrate the In C2 domains such as tower air traffic and systems engineering. The aim of CSE and Leong, 2007). Networked C2 systems CSE approach. control, human cognition is affected is to ensure that technological systems are have paved the way for pervasive and near significantly by naturalistic challenges as designed adequately to support human real-time access to massive amounts of data Cognitive Challenges in well as the level of human expertise and performance in naturalistic work from the battlefield and other information Complex Command and Control experience in the tasks performed. environments. sources from the military enterprise. However, Operational Environments Macrocognition is a collective term used this rapid and easy access to information has frequently to describe cognitive processes The subsequent sections of this article increased the tempo of military operations and functions in naturalistic work settings describe the CSE approach taken by a C2 The definition of C2 from a human-centred and created new challenges to manage and (see Figure 2). These macrocognitive processes systems development team from DSTA to perspective is “a complex, ill-defined sieve out critical information. and functions are influenced by the context develop a C2 system prototype for tower dynamic human decision making process of the operational environment (Klein et al., controllers working in the ATCT. that establishes the common intent and Cognition refers to the way humans think 2003; Cacciabue and Hollnagel, 1995). i.e. how we perceive, learn and reason. transforms that common intent into a In the information-rich and dynamic C2 co-ordinated action” (Pigeau, 1998). environment, human operators have to focus Boyd’s Observe-Orient-Decide-Act (OODA) on multiple tasks simultaneously, and this loop in Figure 1 captures the cyclic suite of Sensemaking allows people to diagnose how the Planning is the process of modifying action current situation came about and to anticipate to transform a current state into a desired could result in cognitive overload. Cognitive how the situation will develop in the future future state. Replanning is modifying, through the deliberate and conscious process of adjusting or replacing a plan that has already overload is a well-known bottleneck in fitting data into a frame of understanding. been implemented. human information processing and originates from limitations in human attention (de Klein’s recognition-primed decision (RPD) model (Klein, Coordination is the way team Greef and Arciszewski, 2008). The modern C2 1998) describes how people members orchestrate the typically rely on their experience sequencing of their actions to command post offers opportunities for the to identify a plausible course perform a task. Teams may benefit design and development of a wide variety of action in challenging and from having resources available complex naturalistic settings at their disposal, knowledge to of decision aids and automation to augment in the field, and use mental leverage and a wider span of task the cognitive work performance of human simulation to evaluate the course control. of action without having to operators and command teams. compare it with others. Maintaining Common Ground However, studies and real-life incidents Attention management is the use refers to the continuous of perceptual filters to determine maintenance and repair have highlighted several unintended the information that a person of calibrated understanding will seek and notice. Attention among members in a team. This consequences of technology on human work management is important as is necessary for coordination, performance. Some of the common problems Figure 1. The OODA loop information technology increases communication and effective the flow of data dramatically. teamwork. in human-systems integration include of C2 activities (Source: Paradis, Breton and Roy, 1999) unbalanced workload, reduced situational Managing uncertainty is important in ill-structured The ability to spot potential problems early is critical awareness, decision biases, mistrust, over- and ill-defined domains. Uncertainty refers to the state in most naturalistic settings. Problem detection cognitive activities in C2. These activities or feeling in which we do not know or understand but depends on the simultaneous reframing i.e. reliance and complacency (Parasuraman feel the need to do so. Uncertainty can arise when re-interpretation of the understanding of the and Riley, 1997). The pitfalls associated with involve the perception of the environment critical data are missing or unreliable, when goals are situation, triggered by the cues received. Skilled unclear, when the problems are not stated clearly, or decision makers can recognise anomalies quickly to relying on technology to solve problems (Observe), assessment of the situation when we are unsure about what to do next. avoid or deflect the consequences. in human cognitive performance have (Orient), decision making over a course sparked growing interest in applications of action (Decide) and implementation of Figure 2. Macrocognition functions and processes (Source: Adapted from Klein et al., 2003) of Cognitive Systems Engineering (CSE) in the chosen plan (Act). However, in stark 29 C2 domains, leveraging our knowledge contrast to the simple and straightforward of human cognitive, perceptual and OODA loop, the C2 domain is naturalistic, Cognitive Systems Engineering Approach to Developing Command and Control Systems 30

operations. The tower controllers have to be After a preliminary assessment of the levels increased significantly when DESIGNING A COGNITIVE proactive in detecting these problems early potential cognitive challenges in ATCT anomalies or unexpected circumstances COMMAND AND and prevent them from severely affecting operations, the development team believed occurred. Therefore, the project team also CONTROL SYSTEM FOR tower air traffic control operations. All these that the C2 system prototype can be designed decided that the VAPS system should focus activities require tower controllers to be very with software assistants to help reduce on assisting the tower controllers when TOWER CONTROLLERS focused and vigilant. the tower controllers’ workload and unexpected situations occur. facilitate reasonable levels of operational An Overview of Air Traffic The Decision-Centred Design performance even under chaotic situations. Phase 2: Knowledge Control Tower Operations These software assistants comprise decision Methodology Elicitation – Uncovering support technologies that are categorised Macrocognitive Challenges The ATCT is a tall windowed structure that into two functional groups: Virtual Assistant The team adopted the Decision-Centred is situated prominently in every airport. A (VA) and Persistent Sentinels (PS). Design (DCD) methodology (Crandall, Klein A critical aspect of the DCD methodology is team of tower controllers in the ATCT is and Hoffman, 2006) in designing the C2 the Knowledge Elicitation phase. Cognitive responsible for managing ground traffic VA are software-based ‘helpers’ that aid system prototype. This five-phase design systems engineers conduct cognitive task around the runways and airborne traffic in human operators by performing routine methodology (see Figure 3) guides the CSE analysis (CTA) interviews to elicit cognitive the immediate vicinity of the airport. tasks such as information retrieval and simple development process to ensure that design difficulties faced by human operators, calculations on their behalf. They may also features of the C2 sytem effectively address potential errors made by novice operators, as The tower controllers’ primary function is support more complex planning tasks such macrocognitive challenges faced by tower well as the knowledge and strategies subject to ensure the timely departure and arrival as impact analysis and resource optimisation. controllers. matter experts use at work. of aircraft on runways under their control. PS are a specific form of software-based Tower controllers also face the constant ‘helpers’ that play the role of ‘watchdog’ An overview of the tasks carried out at each The project team conducted several CTA challenge of ensuring that the aircraft and by offloading monitoring tasks that would phase is as follows. interviews with tower controllers, air traffic passenger safety are not compromised. otherwise require constant human attention. control trainers and tower supervisors. PS allow human operators to divert their Through this process, the team gathered To carry out their responsibilities, the tower Phase 1: Preparation – attention from low-level monitoring tasks information on critical incidents that served controllers must scan the runway and track Understanding the Context to more complex cognitive tasks that cannot as the base data for the Analysis and the positions of airborne aircraft in the be automated easily. The human operators Representation phase. Three scenarios that surrounding air space. The tower controllers The Preparation phase involves are alerted by the PS when unusual events contain good examples of key cognitive have to communicate and coordinate with understanding the domain and nature of happen. challenges faced by the tower controllers many parties including team members, pilots, the tower controllers’ work, as well as were also identified. These scenarios ground vehicle drivers and Terminal Radar the tasks and functions they perform. The The design goal of a VA and PS system (VAPS) were used during the Evaluation phase to Approach controllers on aircraft arrivals purpose of this phase is to search for high for the ATCT is to augment the tower demonstrate the features of the VAPS system and departures. Occasionally, unexpected payoff areas where the CSE efforts should controllers’ cognitive performance in to the tower controllers. events may happen during normal airport focus on. their work environment by addressing the cognitive challenges in maintaining comprehensive situational awareness and Phase 3: Analysis and decision making. The VAPS system functions Representation – Framing as a virtual team-mate that is context-sensitive Macrocognitive Requirements and works together with human operators to achieve operational goals. During the Analysis and Representation phase, the project team collated key findings As tower controllers work in a highly visual from the Knowledge Elicitation phase in a environment, the project team decided that Macrocognitive Requirements Table (MRT). a key objective for the VAPS system is to Macrocognitive challenges are broken down enhance the tower controllers’ ability to into smaller and distinct challenges, each understand situations around the control characterised by critical sensemaking cues tower quickly. After several visits to the and decision anchors. ATCT, the project team understood that 31 many of the controllers’ primary tasks The critical cues and anchors identified in seemed to be procedural in nature and the MRT were especially important in the required little skilled judgement. However, Application Design phase, as they eventually Figure 3: DCD Methodology (Source: Adapted from Crandall, Klein and Hoffman, 2006) the tower controllers’ stress and workload evolved to be key design elements in user Cognitive Systems Engineering Approach to Developing Command and Control Systems 32

interfaces. An excerpt of the table prepared obtained from ground personnel to by the project team is shown in Table 1. determine the type and magnitude of the incident. Difficulties in acquiring Phase 4: Application sufficient information impeded their Design – Transforming ability to perform effective sensemaking. Macrocognitive Requirements To overcome this challenge, one of the features of the VAPS system is the control of into Design Elements pan-tilt-zoom cameras that allows the tower controllers to obtain a close-up video of In this phase, the team designed the visually obstructed areas. features of the VAPS system based on the macrocognitive requirements identified A straightforward solution would be to at the Analysis and Representation phase. provide the tower controllers with manual The following three examples show how control of the cameras. However, the project macrocognitive challenges have led to the team realised that having to control and development of design concepts. monitor the cameras individually would increase the tower controllers’ regular Macrocognitive Challenge 1: workload. The team also observed that Determining Type and Magnitude the various locations and directions these of an Incident cameras were facing affected one’s ability Figure 4. VAPS to overcome macrocognitive challenges in air traffic control to manually correlate the video to a location A recurrent cognitive challenge for tower – this could further exacerbate the tower help the tower controllers quickly associate accurate information and lessens the impact controllers is the difficulties in interpreting controllers’ mental workload. the video to corresponding locations of of human-to-human miscommunication. and understanding incidents as they unfold. the cameras (see Figure 4). Manual control The tower controllers’ direct line of sight may The team decided that the graphical user of each camera may be an additional Macrocognitive Challenge 2: sometimes be obstructed by surrounding interface of the system prototype should burden for tower controllers. Thus, the Determining Courses of Action of buildings and foliage. In these cases, include a two-dimensional (2D) schematic project team developed software agents an Incident they may have to rely on information plan of the work area with reference lines to that will automatically select the most suitable camera to view the corresponding Once an incident has been identified, the Macrocognitive Macrocognitive Why Difficult? Critical Cues/ Potential Design Ideas Macrocognitive ground location when the tower controllers tower controllers have to determine a Challenge Function/Processes Anchors Errors Functions/Processes Involved Addressed click on the 2D schematic plan. This click-and- method to resolve the incident quickly. 1. Determining • Problem • Unable to have • Loud sound of • Delay in • Cameras to • Problem view feature offers an easy and interactive The standard operating procedures (SOP) incident detection a clear visual two aircraft informing enable the ATC detection way for tower controllers to monitor happening • Managing due to Air colliding emergency to look at what • Managing for tower controllers contain checklists of in parking uncertainty Traffic Control’s • Smoke from services due is happening in uncertainty incidents. It alleviates their mental workload actions for many possible events. The VAPS areas • Sensemaking (ATC) distance engines to uncertainty areas that are • Sensemaking (CTA 1) from the • Landing of about magnitude not visible to as they do not have to remember which system is designed with a Course of Action parking areas foreign aircraft of situation the naked eye and through cameras can be used to view the location of • No direct • Trade-space (COA) suggestion feature to automatically binoculars communication between the an incident. The agent-based camera system provide suggestions of relevant key with the foreign safety and • Controls to marshaller who time (anchors) facilitate ATC’s also reduces the tower controllers’ reliance actions according to the SOP checklists (see was directing manipulation the aircraft of cameras on the ability of ground personnel to report Figure 5). 2. Determining • Planning • Uncertain • Magnitude of • May estimate • Provide the ATC • Planning how best to • Coordination about length of incident wrongly the with a way to • Coordination manage • Managing time required (anchor) time required for easily obtain • Managing incoming air attention to resolve • Verbal incident to be and keep track attention traffic in the the incident/ reporting by resolved of the fuel levels event of obstruction pilot of fuel • May carry out of aircraft runway • Unaware of the endurance actions that • Provide the ATC obstruction remaining (usually in are more with a way (CTA 2) endurance of terms of conservative or to monitor the the incoming number of riskier than status of incident aircraft minutes necessary (e.g. resolution • Unaware of remaining) divert all aircraft, process available • Future expected tell low fuel capacity of load aircraft to other airports continue circling) (where low fuel 33 aircraft could potentially be diverted to) Figure 5. A mock-up of the system interface showing COA suggestions Table 1. Selected information from the MRT for the VAPS system (Source: Adapted from Klein and Hutton, 2007) Cognitive Systems Engineering Approach to Developing Command and Control Systems 34

Locations around the control tower are as a means of preliminary evaluation of the Long, W. and Cox, D.A. 2007. Indicators tagged with priority values that reflect system features. The team also used Cognitive REFERENCES for Identifying Systems that Hinder their relative importance with respect to air Indicators (Long and Cox, 2007) which are Cognitive Performance. Paper presented traffic control. The VAPS system prioritises specialised heuristics for evaluating how Cacciabue, P.C. and Hollnagel, E. 1995. at the Eighth International Conference on COA suggestions based on the priority value system features in the VAPS system support Simulation of Cognition: Applications. Naturalistic Decision Making, Asilomar, assigned to the locations where the incident or hinder cognitive performance. Expertise and Technology: Cognition and California, 4-6 June. http://bss.sfsu.edu/ occurred. The COA suggestions are also Human-computer Cooperation: 55-67. New kmosier/NDM8_Proceedings.pdf (accessed contextualised taking into account other A proof-of-concept trial was carried out to York: Lawrence Erlbaum Associates. 15 September 2008) factors such as locations of aircraft on the assess the overall effectiveness of the VAPS ground and runway availability. system. Within a controlled laboratory Crandall, B., Klein, G. and Hoffman, R.R. Paradis, S., Breton, R. and Roy, J. 1999. environment, the team designed experiments 2006. Working Minds: A Practitioner’s Guide Data Fusion in Support of Dynamic Human Macrocognitive Challenge 3: to determine the extent to which the VAPS to Cognitive Task Analysis. Cambridge, Decision Making. Proceedings of the Second Determining the Impact of an system enhanced the tower controllers’ Massachusetts: The MIT Press. International Conference on Information Incident reaction time, quality of decision making Fusion, July 6-8, Sunnyvale Hilton Inn, and quality of awareness. It was found de Greef, T. and Arciszewski, H. 2008. Sunnyvale, California. International Society Combining Adaptive Automation and During the CTA interviews, subject matter that tower controllers using the prototype of Information Fusion. Adaptive Teams in a Naval Command Centre. experts in the tower air traffic control performed better under heavy air traffic Proceedings of the 15th European Conference domain described how unfolding incidents volume conditions and had increased Parasuraman, R. and Riley, V. 1997. Humans on Cognitive Ergonomics: the Ergonomics of affected their ability to guide the take-offs incident management capacity. The tower and Automation: Use, Misuse, Disuse, Abuse. Cool Interaction: 1-4. New York: ACM. http:// and landings of aircraft during operations. controllers also commented that the system Human Factors, 39:230-253. portal.acm.org/citation.cfm?id=1473047 Tower controllers often require substantial prototype helped to improve their situational (accessed 15 September 2008) effort in coordination, decision making and awareness. Pigeau, R. 1998. The Human in Command. sensemaking to resolve an incident quickly. Defence Science & Technology, Issues #2. Hoffman, R.R., Klein, G. and Laughery, The resulting cognitive overload could cause CONCLUSION K.R. 2002. The State of Cognitive Systems tower controllers to overlook subtle but Yeoh, L.W., Mui, W.K. and Leong, S.C. 2007. Engineering. IEEE Intelligent Systems, 17(1): critical considerations that might impact The development of the VAPS system Application of Cognitive Science to Design- 73-75. http://www.computer.org/portal/ tower air traffic operations significantly. demonstrated the effectiveness of CSE in ing Command and Control Systems. Paper web/csdl/doi/10.1109/5254.988462 (accessed designing complex C2 systems to support presented at the Eighth International Confer- 15 September 2008) Since the tower controllers’ ability to guide cognitive work performance. The DCD ence on Naturalistic Decision Making, Asilo- the take-offs and landings of aircraft depends methodology provided a structured process mar, California, 4-6 June. http://bss.sfsu.edu/ Klein, G. 1998. Sources of Power: How People heavily on the availability of runways and to elicit cognitive requirements and translate kmosier/NDM8_Proceedings.pdf (accessed 15 Make Decisions. Cambridge, Massachusetts: taxiways, the project team incorporated a these requirements into system features. September 2008) The MIT Press. Conflict Management module in the VAPS system that informs tower controllers of Apart from its use in designing C2 systems for Klein, G. and Hutton, R. 2007. Overview of impending runway usage conflicts. The single human operators, the CSE approach Sensemaking. Presentation at a Cognitive Conflict Management module incorporates can also be applied in other areas such as Systems Engineering Workshop, Singapore, air traffic control heuristics obtained during human-systems integration in unmanned 23-28 March. the CTA interviews. When a runway is vehicle operations, as well as collaboration unexpectedly occupied by an aircraft or a and workload management in command Klein, G., Ross, K.G., Moon, B.M., Klein, ground vehicle, the Conflict Management teams. D.E., Hoffman, R.R. and Hollnagel, E. module uses a flight information system 2003. Macrocognition. IEEE Intelligent to retrieve pre-defined flight plans Designers of future military systems can Systems, 18(3), 81-85. http://ieeexplore. automatically. The list of aircraft scheduled consider adopting a CSE approach to systems ieee.org/iel5/5254/27031/01200735. to arrive or depart in the near future helps development whenever it is necessary to pdf?arnumber=1200735 (accessed 15 tower controllers to anticipate possible ameliorate cognitive challenges in complex September 2008) delays to the flight plans. work domains. The CSE approach helps system developers design appropriate Phase 5: Evaluation decision aids and other forms of automation 35 technologies to improve system effectiveness, After the prototype was created, the project as well as enhance safety and work team conducted a Cognitive Wall Walk with productivity. tower controllers (Klein and Hutton, 2007) Cognitive Systems Engineering Approach to Developing Command and Control Systems 36

BIOGRAPHY Linus Low Kar Seng was a Senior Engineer (C4I Development). His areas of work involved the design and development of Command and Control Yeoh Lean Weng is Head of the US Defence Technology Office. He is systems for sensemaking, decision making and collaboration. He was also a responsible for prospecting technologies and establishing technology member of the team which spearheaded the development of the Cognitive collaborations with US government agencies, academic institutions and Systems Engineering technical competency in DSTA. Linus was a member innovative companies. He facilitates export control and the release of critical of the Singapore Armed Forces Centre for Military Experimentation Team technologies to Singapore. Prior to assuming this position in July 2010, which won the Defence Technology Prize Team (Engineering) Award in 2005. Lean Weng was Director (DSTA Masterplanning and Systems Architecting) A recipient of the Defence Technology Training Award (Local), he obtained and led the development of new operational concepts and architectural a Bachelor of Engineering (Electrical Engineering) degree with First Class frameworks. Lean Weng has extensive experience working on large-scale Honours, and a Minor in Management of Information Technology from NUS defence engineering systems. As a systems architect, he played a key role in 2003. in developing the Enterprise System-of-Systems Architecture for defence applications. He also developed the systems architecting methodology for masterplanning and transformation. Lean Weng is an adjunct professor at the National University of Singapore (NUS). He is the President of International Teh Shi-Hua is a Senior Engineer (C4I Development). She is currently Council on Systems Engineering (INCOSE) Singapore Chapter, the INCOSE managing a project to develop a Tower Air Traffic Command and Control Region VI Representative to the Member Board and an INCOSE Fellow. He system. She worked on the development of the Virtual Assistants and is the Vice President and Chairman of the Industrial Group in the Institution Persistent Sentinels subsystem as well as a prototype for an information of Engineers, Singapore (IES). He is a founding member of the IES Certified management system. Under the DSTA Undergraduate Scholarship, Shi- Systems Engineering Professional and the Certified Project Manager Registry. Hua graduated with a Bachelor of Engineering (Electrical Engineering and Lean Weng received his Bachelor and Master of Science degrees from NUS in Computer Science) degree from the University of California-Berkeley, US 1983 and 1987 respectively. He further obtained two Master degrees in 1990 in 2005. She further obtained a Master of Science (Management Science and a Doctor of Philosophy (Electrical Engineering) degree from the Naval and Engineering) degree from Stanford University, US in 2006. Postgraduate School (NPS), US in 1997.

Oliver Tan Kok Soon is a Principal Engineer (C4I Development) and has many years of experience in developing C4I systems for the Singapore Armed Forces (SAF). His current areas of work include cognitive systems engineering and decision-support systems. Oliver was a member of the Enterprise Command and Control Information System Team which clinched the Defence Technology Prize Team (Engineering) Award in 2003. He obtained his Master of Technology (Knowledge Engineering) degree from NUS in 2001 where he graduated as the top student. A recipient of the DSTA Postgraduate Scholarship, he further obtained a Master of Science (Modelling, Virtual Environments and Simulation) degree from NPS, US in 2005 where he received awards for his outstanding thesis on surface warfare and for graduating at the top of his class. Oliver is currently pursuing a Master of Science (Human Factors Engineering) degree from Nanyang Technological University.

37 ABSTRACT Enabling Army Learning The fast-changing operating landscape and undefined threat scenarios require Third Generation soldiers to adapt, learn and operate rapidly in a wide spectrum of operations. In recognition Transformation of this, the Army’s Training Transformation Master Plan (T2MP) was developed with the vision of producing leaders and thinking through Infocomm Technology soldiers. The transformation entails a fundamental pedagogical paradigm shift – from passive and one-dimensional instructor- led lessons to a more collaborative, immersive and self-directed learning environment. Under this masterplan, one of the key strategic thrusts is to leverage technology as a vehicle to enhance and innovate the way institutes teach and how soldiers learn.

The LEARNet programme, as part of T2MP, is the keystone to a Learning Army. It strives to deliver the holistic learning system in three ways: transformation of curriculum, deployment of immersive rich media content and mobilisation of knowledge through a learning network and portal. It enables the Singapore Armed Forces (SAF) to keep pace with revolutions in the use of social and rich media applications in learning. It also allows learners to tap familiar and effective pedagogies to quickly assimilate new knowledge and skills.

However, media-heavy applications are typically bandwidth-intensive. Existing networks can be easily stretched beyond their capacities. This emerged as one of the key challenges for the programme. This article describes how Content Distribution Network (CDN) technology was successfully deployed during the LEARNet Phase 1 implementation to overcome the bandwidth challenge. This is a major breakthrough as CDN implementation is a pioneer technology for the Ministry of Defence and the SAF.

Lee Hwee Ling Darren Teo Wee Hong Cindy Seah Ye Leng Enabling Army Learning Transformation 40 through Infocomm Technology

One critical success factor is the ability to INTRODUCTION infuse technology possibilities into the SAF’s core learning process and training The learning culture in the Third Generation curriculum. To support this transformation (3rd Gen) Army must keep pace with initiative, one of the strategic approaches increasingly challenging operational is to harness technology, engineering requirements, evolving technological and learning science resources to foster a landscape and changing user expectations learning environment of creativity and innovation. In this technology-enabled world Generation Y (Gen Y) soldiers who are of learning, instructors and trainees are able technologically savvy and comfortable with to collaborate and co-create knowledge, Infocomm Technology enabled training to produce bigger and better ideas as a make up the demographic profile of the collective whole. Army today. Thus, a fundamental change in learning pedagogical approaches and The LEARNet learning system can be knowledge delivery methods is necessary for conceptualised as an end-to-end continuum effective engagement of these digital age that spans across front, middle and back learners. office operations, with key components as shown in Figure 1. In line with the Ministry of Education’s (MOE) `Teach Less, Learn More’ and Infocomm Front Office Development Authority of Singapore’s (IDA) Figure 1. Holistic LEARNet learning system iN2015 initiatives, the current educational Enhanced Learning Space and climate is conducive for the Army to Knowledge Mobility of time and location. The network capability the learning experience of the trainees and transform its learning and training landscape enables trainees to read content and engage enhance their ability to accomplish assigned as well. The LEARNet strategy is to tap rapid in collaborative learning while instructors missions for the organisation. innovations in consumer technologies such can deliver content on a just-in-time basis as THE LEARNET INITIATIVE as ‘iPAD-like’ mobile devices to offer familiar they take on the role of online coaches and Lifelong Learning – A HOLISTIC LEARNING and intuitive front-end interfaces to the subject matter experts. LEARNet engenders a spirit of lifelong SYSTEM learners. The initiative equips each trainee and instructor with a personal mobile Middle Office learning as learners are provided with a computing device. Having such a device not personal Learning Account to track their In 2008, the Army initiated a new wave of only exposes the trainee to lessons beyond Personal Learning Portal professional learning experience throughout training transformation. LEARNet, as part classroom settings, it also provides a window their service terms. This Learning Account of the larger Army Training Transformation to the entire system. The trainee gains The portal aims to create a 3rd Gen also allows learners to consolidate their Master Plan (T2MP), was identified as the access to the repertoire of content as well as Professional Learning Space for individuals competency records, journals and reflections. keystone to a Learning Army. DSTA, as learning and training administration and teams to come together virtually to learn Thus, the SAF can assess the learning progress the enterprise architect and programme activities. Trainees are able to carry out self- and build knowledge at their preferred pace of its personnel and identify any competency management office, was entrusted with the directed learning activities, conduct personal and convenience. Social media and Web 2.0 gaps within the institute. task of ensuring the successful delivery of the reflections and peer appraisals, perform technologies connect people, knowledge LEARNet programme. self-evaluation, and retrieve checklists for networks and expert systems, allowing the Immersive Rich Media and Rapid different types of operations during training. efficient harnessing of intellectual capital and Content Creation A rapid pilot implementation of the LEARNet collective knowledge. Through the portal, initiative was carried out at the Basic Military To enable the exchange of knowledge the learner gains better control of his self- Unlike traditional teaching methods, rich Training Centre (BMTC), Officer Cadet School, between trainees and instructors, a wireless development by building his own learning media resources enable better engagement, Specialist Cadet School and Signal Institute network infrastructure links up the entire process according to his personal aspirations visualisation and deeper understanding of the in Phase 1. The first phase was completed training institute so that learners can have and competency gaps. This helps to improve successfully, while the remaining Singapore access to the knowledge network regardless Armed Forces (SAF) training institutes and 41 units will be covered in Phases 2 and 3. Enabling Army Learning Transformation through Infocomm Technology 42

can take a long time to download. This affects depicts the CDN architecture design in the the system’s performance and the users’ iNET environment of the Ministry of Defence experience. During Phase 1 implementation, (MINDEF) and the SAF. the project team realised that the constant pressure and speed of expanding the network The new CDN system had to be integrated bandwidth can never catch up with the fast- within the iNET environment as part of a growing demand and proliferation of rich larger complex system. The team considered media and high-bandwidth applications. the following key design principles:

OVERCOMING THE Scalability CHALLENGE The system needs to cater for future growth in user population. The approach To overcome the bandwidth limitation, the was to ensure scalability in managing the project team evaluated several alternatives infrastructure instead of having to expand and also studied similar initiatives the network bandwidth constantly. Phase 1 implemented elsewhere, such as in MOE, of the implementation was targeted at four IDA and academic institutions. Eventually, training institutes serving 10,500 learners a content distribution network (CDN) was in total. The next phase will reach out to assessed to be the best solution. another 240,000 active National Servicemen (NSmen). Thus, the architectural design must Figure 2. Wide range of interactive rich media in LEARNet A CDN is a network of systems in which be modular and extendable while being easy multiple copies of data are placed at various to configure, maintain and deploy. hubs to reduce bandwidth requirements subject matter (see Figure 2). It also increases Change Management and maximise speed of access to the data training realism. To meet the operational within the Wide Area Network (WAN). Performance tempo, just-in-time learning has to be It is critical that commanders, instructors and Future requests for a particular content will supported by rapid content development trainees understand the strategic intent of be served from the closest hub, without The system response time determines the capability. Rapid content development the learning transformation and are familiar the need to gain access to the source at the quality of videos streamed to viewers. After technology, augmented by additional with the new technological tools. The project central site. This mechanism leads to a more benchmarking best practices, the team manpower from an integrated workforce, team introduced a customised change direct, and hence faster, delivery. Figure 3 targeted a response time of not more than allows instructors to deliver and deploy high management programme to encourage impact learning content that can be shared trainers and learners to be receptive to across all Services within three to four hours, the system. The programme also ensures a as compared to four to six weeks in the past. smooth operationalisation of the learning system. Back Office Governance Curriculum Transformation and Pedagogical Paradigm Shifts The central governing agency sets directions, formulates policies, manages risks and The instructional mode will shift away from ensures a consistent implementation of passive, uni-directional and instructor- training policies and syllabi in the various led lectures to a more immersive and training institutes. collaborative self-directed learning environment. There is an ongoing comprehensive curriculum review to identify THE CHALLENGE and restructure courses to the new mode of delivery and engagement. The delivery of rich media over a bandwidth- 43 limited network that connects the front, middle and back offices is challenging as such content is usually bandwidth-intensive and

Figure 3. System architecture of the CDN in MINDEF Enabling Army Learning Transformation through Infocomm Technology 44

eight seconds for 90% of the hits during peak for deployment. The ability to optimise these A good understanding of the usage pattern Building the Right Model of periods. The quality of the videos had to be factors reduces the need for expansion in and content resulted in a combination User Behaviour – Utilisation maintained at a resolution of at least 640 by remote server rooms and the re-equipping of of solutions. The team adopted the same 480 pixels at 512 kilobits per second (kbps), power and cooling systems. This leads to cost principle of CDN but differed slightly in and Traffic Modelling which is better than the quality of typical efficiency with lower upfront investment and the approach for institutes with a large While the issue of scalability was being YouTube videos. less recurrent maintenance. population (i.e. more than 5,000 personnel) managed, the traffic patterns were further such as BMTC. For such institutes, selected analysed. Understanding the traffic content was pre-loaded in the computing Improved System Reliability IMPLEMENTING THE patterns was critical for fine-grained system devices instead of the remote cache. This CONTENT DISTRIBUTION performance tuning. LEARNet shares the WAN with many systems. changed the system usage traffic pattern and The LEARNet implementation should not NETWORK IN MINDEF demand. Through this implementation, the The initial implementation strategy was to result in any degradation of the operations team realised that rich media content varies AND THE SAF deploy LEARNet to the training institutes one and performance of other systems such as and can be handled in the following ways: at a time. However, with further analysis, email and other business applications. Hence, For successful operations-technology (ops- it was noted that the curriculum and usage LEARNet’s design should reduce dependence tech) integration, a good understanding • Pre-loading of content in devices through behaviour varies across different training on the WAN. of the customer’s business is of utmost the use of content push down via system institutes. Hence, the implementation importance. In implementing the CDN, the management tools strategy was reviewed and revised team analysed user content, usage and traffic Security subsequently. The new approach called patterns before designing a system with • Pre-loading the latest copy of content in for progressive deployment in all training adequate scalability and performance-tuning the remote cache at pre-defined times of the Due to security and operational institutes. A subset of military units i.e. to support the anticipated load. The original day (i.e. pre-positioning) requirements, all transmissions across the companies within each training institute was proliferation strategy was also tweaked to network must be encrypted. Secure Sockets chosen for each phase of rollout (see Figure allow a more customised rollout to each • Releasing the latest copy of the content Layer (SSL) encryption is the de facto 4). This offered the team opportunities to training institute to better meet its needs. to the remote cache only upon request by the standard for encrypted and authenticated better understand the training institute so as ‘first user’ (i.e. dynamic caching) communication between end users and to develop more accurate profiles, models of application servers. The caching technology Stretching Beyond Design utilisation and user traffic. selected must thus have the ability to Limits optimise encryption algorithms and reduce SSL handshakes over the WAN. Through a paper design exercise, the team October 2009 August 2010 assessed that the current network can Interoperability and Open neither support the desired load nor meet Sequential proliferation to Training Institutes (TI) one at a time Standards the performance objective. Expanding the network would incur significant time and Full Scale Implementation at TI 1 Any technology introduced within the hefty monetary investments. Full Scale Implementation at TI 2 environment of MINDEF and the SAF has to

be integrated with existing legacy systems Further analysis of the anticipated usage Initial Strategy Full Scale Implementation at TI 3 and across multiple networks. It also has to be pattern offered a better understanding of Full Scale Implementation at TI 4 flexible enough to cater for interoperability dealing with the challenge. It was noted that with future systems. the surge in traffic was due to two factors: (a) trainees logging on to follow the course Progressive proliferation to subset of TI To support interoperability, the system was content during lecture-style lessons and designed based on industry standards for file (b) learners from the BMTC using the system formats, external protocols and interfaces. in allocated time slots. It was observed that 4 Companies within TI 1 4 Companies within TI 1 4 Companies within TI 1 instructors on the ground had allocated fixed The technologies are also platform 4 Companies within TI 2 4 Companies within TI 2 4 Companies within TI 2 independent to prevent reliance on a single time slots for trainees to be engaged in a 4 Companies within TI 3 4 Companies within TI 3 4 Companies within TI 3 vendor or service provider. self-directed mode for some of the courses, Revised Strategy to better control the progress of the overall 3 Companies within TI 4 3 Companies within TI 4 3 Companies within TI 4 curriculum. This case illustrated that time and 45 Total Cost of Ownership Figure 4. Implementation strategy effort are required to achieve the desired learning transformation which involves The amount of hardware, rack space and mindset and behavioural changes. power requirements were key considerations Enabling Army Learning Transformation through Infocomm Technology 46

Assurance and Confidence are emerging and constantly improving in the video stream and achieves a faster response WAY AHEAD consumer domain. With these developments, time. The consumer domain also tends to see It was important to validate the paper it would be possible to detect a user’s shorter product cycles and rapid innovations. The next phase of LEARNet programme design and tests in a real world environment available bandwidth and central processing Thus, constant monitoring and importing of will continue to develop the close ops-tech to ensure that performance specifications unit capacity in real time – this allows an such innovations will remain critical. integration and strong working relationship can be met. User confidence in the system automatic adjustment of the quality of a with the SAF. Future demands will continue was critical in managing the change in the to pose challenges and create opportunities LEARNet project. in new areas such as: BIOGRAPHY The challenge was to be able to generate • Application to other domains – The system loads in the actual infrastructure so implementation has shown that in spite that bottlenecks could be identified and Lee Hwee Ling is a Principal Engineer (Enterprise IT). She is instrumental of limited network bandwidth, effective eliminated. The system could also be fine- in conceptualising and developing the master plan for the Learning delivery of rich media over the network tuned for better performance before trainees Management programme of the Singapore Armed Forces (SAF). She can still be achieved through innovative use it. An end-to-end performance testing is overseeing the development and implementation of the LEARNet technology and proper content governance. exercise was conducted. This involved the programme to enable self-directed learning in the SAF. She was also actively The CDN concept and technology can also use of load generation software to simulate involved in the SAF’s Knowledge Management and Leadership Competency be extended to our secure network to meet actual trainee behaviour and the deployment Development initiatives. She was a member of the Knowledge Management the operational demands of other business of multiple monitoring agents at all network project team that won the Platinum Knowledge Management Excellence domain areas such as those of intelligence traffic routes. Expertise and resources from Awards in 2010. Hwee Ling obtained a Bachelor of Computing (Information and medicine. various teams were employed in the set-up Systems) degree with Honours from the National University of Singapore and configuration of components ranging (NUS) under the Defence Technology Training Award. She also obtained the • Extending the Reach – It is also from network, firewall, caching database, Senior IT Project Management certification in 2010. intended that the next phase reach out to application server and proxies. all training institutes, schools and possibly all units within the SAF. In addition, access will The exercise was completed after a month also be extended to home users by tapping of intense and thorough testing and fine- on the Next Generation National Broadband Darren Teo Wee Hong is a Senior Engineer (Enterprise IT). He is the Technical tuning. The team had a much deeper Network to connect NSmen and other Architect for the LEARNet programme and manages the implementation understanding of the system and could react external educational institutions and armed of LEARNet for SAF training institutes and units. He leads the front-end quickly to unexpected problems when the forces. planning exercise and develops the enterprise system architecture, working system was rolled out. A summary of the test on aspects that include content distribution network deployment, last mile results in Table 1 demonstrates significant • Emerging Technologies – New infrastructure implementation, integrated workforce and computing devices improvements with the CDN strategy. technologies for delivering high-definition proliferation, and content development. He was also involved in knowledge videos such as adaptive bit rate streaming management projects where he conceptualised and implemented common applications for the DSTA intranet. He was a member of the Knowledge Management project team which won the Knowledge Management Response Time Performance Excellence Awards in 2010. Darren graduated from NUS with a Bachelor of (90% percentile) Improvement Computing degree in 2001.

Login to typical portal web page

Without going through cache 5.87s Cindy Seah Ye Leng is a Senior Engineer (Enterprise IT). She is with the 38% Learning Management Systems team managing eLearning projects for the Going through cache 3.63s SAF. Previously with the DSTA Business Management Information System team, Cindy worked on several business system projects including the Streaming an 11 megabyte video content through the video streaming application blueprinting and implementation of the Procurement System which won the SAP Overall Best Implementation Award. Cindy graduated with a Master Without going through cache appliance 30.20s of Commerce (eCommerce) degree from the University of Queensland, 47 86% Australia. Going through cache appliance 4.00s

Table 1. Test results Introduction to ABSTRACT The submarine is a unique platform that is capable of dealing with conventional as well as asymmetric threats from the Submarine Design littorals. As a stealthy platform that possesses a robust capability for conventional open water anti-surface and anti- submarine warfare, a submarine is also increasingly called upon to undertake intelligence gathering, counter-terrorism and special force operations. The need for stealth, range and flexibility while controlling the size and cost of the submarine has resulted in innovative submarine designs. Four solutions are presented in this article.

Ong Li Koon Liu Chee Kong Toh Chee Wee

Introduction to Submarine Design 50

component and system levels without INTRODUCTION the need for a substantial growth in the submarine’s size. The modularity concept Since the end of the Cold War, navies have Control Living is embraced at all levels of submarine 2 Engines AIP Room Quarters shifted their focus from preparing for open- production and operation, namely: water conflicts to dealing with conventional as well as asymmetric threats from the littorals. • Modularity at Design At the same time, navies have to fulfil new (Deconittignies, 2001) requirements such as supporting intelligence • Modularity at Construction Control Living gathering, counter-terrorism and special 2 Engines 2 Engines AIP (Deconittignies, 2001) Room Quarters force operations. Submarines offer a unique capability proposition as a stealthy platform, • Deployment of Modular Payloads and they also possess a robust capability to Figure 2. Modular customisation for customer’s needs meet both the conventional demands of Modularity at Design open water warfare and new demands of or a shorter one for a dedicated role. Figure target acquisition systems; special operations navies. However, acquiring and maintaining Modularity can be applied to the design at 2 shows a generic submarine type in two forces (SOF) support systems; strike weapons; a submarine fleet is costly. Therefore, the key the component, system and boat levels. variants with different engine modules. mines and mine countermeasures; undersea challenge for submarine designers is to strike It allows tailoring to suit individual communications and sensor network systems; a balance between the need for stealth, range requirements without the need to change and unmanned vehicles. Modularity of and adaptability, and the need to control the the entire design of the submarine. Modularity at Construction payload removes the need to incorporate size and cost of the submarine. a single submarine with the complete The concept of modularity relies on a clear ‘Modularity at Construction’ refers to set of payload options required for all its In response to these challenges, this article segregation of functions in the submarine the sharing of submarine manufacturing capabilities. Instead, the submarine can presents four critical solutions, namely: as illustrated in Figure 1. For example, a processes at different shipyards. Each hull be designed to allow the customisation submarine can be divided into modules with section can be fabricated and assembled in of payloads depending on specific mission • Modularity sections dedicated to certain functions e.g. shipyards located in different countries. The final assembly is then carried out at one of needs. • Design for Special Operations Forces diesel engine room and living quarters. These modules are designed as stand-alone systems these shipyards. Modularity at Construction • Battery Technology Figure 3 illustrates a modular sail section with their own support functions. avoids production bottlenecks and speeds using the Universal Modular Mast (UMM) • Air Independent Propulsion (AIP) up the manufacturing process considerably. concept, which consists of cartridges installed Some of these modules can be customised Such a concept, however, requires an in the sail of a submarine. These cartridges subsequently to meet the requirements of integrated information system, extensive MODULARITY are used to house different mission-specific different customers. For instance, the diesel quality control and tightly controlled masts as required. Besides masts, auxiliary engine module can be customised to provide production tolerance. This concept has Cutbacks in fleet sizes and budgets have led systems can also be fitted into the cartridge additional power in a long version with four been used by the French submarine builder, to the demand for more capabilities to be sets of diesel engines, or to provide less DCNS, in the manufacture of the Scorpene incorporated in submarines so that they can power in a short version with two sets of family of submarines for Chile and Malaysia perform multiple roles. As a result, submarine diesel engines. Similarly, there can be two (Deconittignies, 2001). Both DCNS and Spain- builders have incorporated modular design versions for the sail section: a longer one for based Navantia share the construction and in their latest generation of submarines. additional masts to meet multiple operations, outfitting of the hull sections. The final Modular design allows the addition of assembly is then carried out at either DCNS capabilities and easier upgrades at the or Navantia.

Modular Payloads

The introduction of modular payloads is one

Diesel Engine Control Living of the most significant and newest changes AIP Room Room Quarters in submarine design. Instead of carrying only 51 a fixed payload of torpedoes, submarines with modular payloads can be re-configured Figure 1. Segregation of functions in a submarine for a range of payload options. These include intelligence, surveillance, reconnaissance and Figure 3. The UMM concept Introduction to Submarine Design

52

of the UMM. For example, a machine gun as heavy as 30 tonnes and very few navies can can be stowed in a UMM cartridge, which can afford submarines that are large enough to then be raised to deploy the gun. The UMM carry them. concept has been adopted by new generation submarines such as the UK Royal Navy’s latest Designers of smaller submarines which are Astute-class submarines. incapable of carrying such heavy locks have explored innovative means to achieve covert DESIGN for SPECIAL delivery. For example, the German builders of the U212A class diesel-electric submarines Figure 5. A re-configurable torpedo room OPERATIONs FORCES overcame the lack of space in the submarines by modifying part of the sail fin to become an Figure 5). This allows the submarine to be Lead-Acid Battery The SOF form an important element of internal lock (Wallner, 2006). customised without having to lengthen its modern warfare. Armed forces require the hull to house the additional SOF crew. The lead-acid battery has been the battery SOF to perform a wide variety of missions Besides supporting the covert delivery and of choice for submarine applications. Despite such as Search and Rescue, Reconnaissance extraction process, the submarine has to The US Virginia-class submarine is an having a low energy-to-weight ratio and a and Sabotage, and Forward Observation. allocate supporting resources to the SOF excellent example of an SOF-friendly correspondingly low energy-to-volume ratio, For these operations, the SOF may be sent to team, such as accommodation, food, stowage submarine. It has a dedicated modular lock lead-acid batteries are able to supply high their mission areas using submarines, aircraft, space for equipment, as well as mission section that can deliver a nine-man SOF surge currents and maintain a large power- helicopters, parachutes or surface craft. As planning and control areas. At the same time, team for a single operation. The submarine to-weight ratio. This explains why they are submarines are the only platform that can the submarine has to maintain sufficient also features on-board modular spaces chosen for submarine applications. The lead- guarantee covert delivery, they are a vital space for its own equipping needs to maintain that can be re-configured to accommodate acid battery stores electrical energy in the element of the SOF. its core fighting capability. It is difficult to additional bunks, stowage space as well as form of chemical energy and releases this optimise the small conventional submarines mission planning and control equipment. The stored energy into an electrical circuit as the Covert Delivery to support SOF operations while maintaining submarine can also carry swimmer delivery battery discharges. The operation of a lead- their core war fighting capability without a vehicles and a decompression chamber acid cell is described in the reversible chemical Up to the 20th century, covert delivery of substantial increase in the submarines’ size. externally (Graves and Whitman, 1999). reaction shown in Figure 6. the SOF could only be achieved by a large submarine with a dedicated external lock Modular Add-on for the SOF BATTERIES FOR that is carried as an attachment on the Operations submarine hull. Figure 4 illustrates an SUBMARINE example of such an arrangement. This lock To overcome size constraints, submarine APPLICATIONS acts as an interface between the submarine designers have used modular design to create and the sea to transfer the SOF in and out capacity to accommodate the requirements Batteries are standard features in all of the submarine while keeping the interior of the SOF without compromising the submarines to provide standby and of the submarine dry. Following the flooding submarine’s performance and core mission propulsion power (Szymborski, 2008). Prior and pressurisation of the lock, the SOF team is capability. For example, the torpedo room to the advent of AIP and nuclear technology, able to swim out to its mission area. The SOF can be designed such that the designated a submarine’s submerged endurance team also returns to the submarine via the torpedo racks can be removed to allow the depended entirely on its battery life. Thus, the lock after its mission. These transfer locks are fitting of additional bunks for the SOF (see time required to charge its batteries remains as one of the submarine’s key performance indicators – this determines how long a submarine has to snorkel and risk detection by adversaries. While lead-acid batteries have been the standard used in submarines, their dominance is increasingly challenged by a Figure 6. Operation of a lead-acid battery new generation of batteries that offer better power and energy density. Figure 4. Covert delivery of the SOF using external lock 53 Introduction to Submarine Design

54

In line with the modularity design concept, AIR INDEPENDENT all AIP systems are built as modular plug- PROPULSION in sections (Bergande and Larsson, 2003). This characteristic makes them suitable for The period of time that a conventional deployment in new and existing submarines. submarine can remain submerged The following sections describe the dominant continuously is limited by its battery AIP systems. capacity. Despite emerging battery technologies, it is difficult to achieve a Stirling Engine quantum leap in the submerged endurance due to volumetric and cost constraints. The Stirling engine solution developed by Submarines with AIP capabilities can extend Kockums is regarded as a well-established AIP their submerged endurance from days to technology. All Swedish submarines employ weeks. the Stirling AIP system. The Japan Maritime Figure 7. The charging and discharging sequence of a Li-ion battery Self Defence Force also deploys Stirling Conventional submarines run on diesel engines in its Soryu-class submarines. engines which require oxygen for Lithium-Ion Battery Major submarine designers such as DCNS combustion. AIP enables the submarine to The Stirling engine produces heat by burning and Howaldtswerke-Deutsche Werft operate without the need to surface or use low sulphur diesel fuel and oxygen (stored in Lithium-ion (Li-ion) batteries are currently (HDW) have initiated research and its snort mast to access atmospheric oxygen. cryogenic tanks) in a pressurised combustion one of the most popular types of battery for development programmes to explore the The oxygen required for combustion is stored chamber. The heat is then transferred to portable electronics. They have a superior replacement of lead-acid batteries with on board as liquid oxygen. AIP technologies the engine’s working gas (usually helium) energy-to-weight ratio and a slow loss of Li-ion batteries. include Stirling engines, fuel cell (FC) systems, operating in a completely closed system. charge when not in use. Lithium is one of the steam turbine systems, and closed cycle The working gas forces the pistons in the lightest metals and has great electrochemical Molten Salt Battery diesel systems. Thus, nuclear propulsion can engine to move, thus producing mechanical potential. In addition to the wide-ranging also be considered as AIP. However, due energy to drive the alternator as illustrated in applications of Li-ion batteries in the The molten salt battery is a class of electric to the size of nuclear submarines and their Figure 8. The combustion pressure is higher consumer electronics domain, there is also cells that uses molten salts as its electrolyte. seemingly limitless submerged endurance, than the surrounding seawater pressure, a growing demand for it in the defence, It offers a higher energy density through a nuclear propulsion is seldom mentioned which allows exhaust products dissolved in automotive, and aerospace industries. proper selection of reactant pairs (i.e. anode under the same category as non-nuclear AIP seawater to be discharged overboard without This is due to the high energy density and or cathode) and better power density by submarines. the use of a compressor. This results in low technological maturity of Li-ion batteries. means of a high conductivity molten salt Figure 7 shows a simplified diagram of the electrolyte. It is used in applications where charging and discharging sequence of a high energy density and power density are Li-ion battery. required.

One of the key advantages of Li-ion batteries The main drawback of the molten salt is their ability to be moulded into different battery is the need to charge the battery shapes and sizes to fill any space available constantly so that the electrolyte will remain in the devices they power efficiently. It has in liquid state and be ready for use when a low self-discharge rate of approximately required. If the battery pack is shut down five to ten percent, which is significantly and left to solidify, a re-heating process must lower than other battery types in the market. be done to restore the battery pack and this No memory and scheduled cycling is needed usually takes three to four days. The most to prolong the battery’s life. Due to these notable example of molten salt batteries desirable traits, Li-ion battery systems were used for underwater application is the tested for application in underwater vehicles molten sodium aluminium chloride based and have demonstrated high potential in ZEBRA battery that was first developed in 55 replacing lead-acid battery systems in diesel- 1985. However, its application in submarines electric submarines. has been limited, with usage seen only in small submersibles rather than submarines. Figure 8. The Stirling AIP system Introduction to Submarine Design

56

Figure 9. The FC AIP system

Figure 10. The MESMA system

infrared signature and noise emission levels. The FC system is operational on board the An established reformer technology will The submarine’s submerged endurance is German U212A class submarines and the resolve the issue regarding the supply and CONCLUSION determined by its storage capacity for the Type 214 class submarines. It has also been storage of hydrogen, which allows the FC AIP liquid oxygen. selected for the S80 class submarines built solution to be implemented more easily in The key challenge to submarine design is to for the Spanish Navy. The FC system is conventional submarine designs. balance the requirements for stealth, range Fuel Cell System advantageous because the only by-product and adaptability with the need to control is pure water and it does not generate any MESMA cost and size of the submarine. This article The FC system is developed by HDW exhaust gas. Furthermore, the FC system is presents critical solutions in response to these modern challenges. in Germany. The system uses Polymer much quieter than other AIP systems, has The French Module d’Energie Sous-Marine Electrolyte Membrane (PEM) fuel cells. the lowest oxygen consumption rate and Autonome (MESMA) closed cycle steam- PEM fuel cells are electrochemical energy potentially offers the highest underwater turbine system burns ethanol and liquid converters in which hydrogen ions and endurance. oxygen in a combustion chamber, generating oxygen ions are combined to produce steam to drive a turbo-electric generator. electrical charge as illustrated in Figure 9. However, the hydrogen required in FC Figure 10 illustrates the MESMA’s working Similar to the Stirling engine, the FC system systems is stored in the form of liquid principle. MESMA is currently fitted in the generates electricity at a slow and steady rate, hydrogen in metal hydride which requires Pakistani Agosta 90B class submarines. suitable for low-speed submarine operations. costly maintenance and support facilities. The For high-surge operations, the submarine S80 class submarines will feature reformer In the search to extend the submerged relies on its regular battery system, which in technology that generates the required endurance and operating capability of turn is recharged by the FC system. hydrogen from ethanol through a reformer. conventional submarines, AIP technologies proved to be a cost-effective solution.

57 Introduction to Submarine Design

58

REFERENCES BIOGRAPHY

Bergande, M. and Larsson, L. 2003. AIP Plug-in Sections: A New Submarine Standard, Naval Ong Li Koon is a Senior Engineer (Naval Systems) managing the submarine Forces Special Issue SUBCON 2003. upgrade programme. He was previously the Deputy Head of the Resident Project Office overseas. He was a member of the Specialised Marine Deconittignies, C. 2001. Submarine Craft Project Team which received the Defence Technology Prize Team Modularity: An Innovative Concept Applied (Engineering) Award in 2006. Li Koon obtained a Bachelor of Engineering to the French Submarines. Paper presented (Naval Architecture and Ocean Engineering) degree with First Class Honours at the Naval Platform Technology Seminar, as well as a Master of Science (Naval Architecture) degree from University Singapore, 10 May. College London, UK in 2000 and 2001 respectively.

Graves, B. and Whitman, E. 1999. The VIRGINIA Class – America’s Next Submarine, Undersea Warfare, Winter 1998/1999. http://www.navy. Liu Chee Kong is a Senior Engineer (Naval Systems). With experience in mil/navydata/cno/n87/usw/winter99/virginia_ managing operations and support requirements for naval weapon systems, class.htm (accessed 25 June 2005) he is currently responsible for setting up the full range of support facilities and infrastructure for naval weapon systems. Chee Kong holds a Bachelor of Szymborski, J. 2008. Lead-Acid Batteries For Use Engineering (Mechanical Engineering) degree from Imperial College London, in Submarine Applications. http://ieeexplore. UK where he obtained awards of the Most Outstanding Student and the ieee.org/iel5/8390/26433/01177196. Greatest Merit in Mechanical Engineering. He further obtained a Master of pdf?arnumber=1177196 (accessed 17 August Science (Mechanical Engineering) degree from Stanford University, US in 2010) 2002.

Wallner, R. 2006. German Submarines – Capabilities and Potential, RUSI Defence System. Royal United Services Institute Defence Systems, Autumn 2006. http:// Toh Chee Wee is an Engineer (Naval Systems). He is currently stationed www.rusi.org/downloads/assets/German_ overseas managing the modification and upgrade works for a naval project. Submarines.pdf (accessed 29 November He was previously involved in the combat system upgrade of a naval 2010) craft project. Chee Wee obtained a Bachelor of Engineering (Mechanical Engineering) degree with Honours from Nanyang Technological University in 2008.

59 ABSTRACT Secure Mobility Mobile technology has advanced significantly since the first commercial cell phone was introduced in 1973. This advancement for the Enterprise of technologies, coupled with the prevalence of the Internet, have opened up new and exciting opportunities for consumers and businesses.

However, the advancement of mobile technologies is predominantly consumer-oriented. For an organisation to use the mobile phone as a corporate productivity tool, the security and manageability gaps must be closed.

This article highlights the key challenges in integrating mobile devices within the defence ecosystem. It discusses critical factors such as security threats, technical constraints, as well as the supportability and maintainability of deploying these mobile devices in an enterprise environment. This article also addresses key issues in the design and development of enterprise mobility solutions.

Adrian Toh Thiam Huat Tan Shyh Hae Lim Dawei Secure Mobility for the Enterprise 62

It is also important to address the generic ENABLING ENTERPRISE challenges of supporting a mobile device. Enterprise Consumer MOBILITY Network stability, different form factors, battery capacity and other factors create Technology Adoption Conservative Aggressive Mobility has come a long way since 1973 an environment that is often beyond the Sensitivity to Cost High Medium when Motorola invented the first commercial enterprise’s control. cellular portable telephone. Approximately Scale of Deployment Large pool of users Single user 793g in weight and 25cm in height, the The arrival of Smartphones led to higher user Motorola DynaTAC 8000X may be considered expectations that cannot be supported by Solution Lifespan Extended Short extremely unwieldy by today’s standards. current technologies. This is a typical phase However, it was the first mobile telephone that happens in the society’s adoption of new Tolerance for Failure Low Medium that could connect to the telephone network technologies, known as the ‘Peak of Inflated without the assistance of a mobile operator. Expectations’ (Milanesi et al., 2009). Main Priorities Security and manageability Design and applications

Therefore, the project management team Today, the availability of high-speed, low- Table 1. Differences between an enterprise and a consumer when adopting a mobile device cost mobile Internet access and related (PMT) has to develop creative solutions, social applications are pivotal driving forces perform systems integration and manage for the mobile device industry. With rapid user expectations driven by rapid Table 1 illustrates the differences between an tedious to ensure that all concerns are duly advancement of technologies, enterprises development in the consumer market. enterprise and a consumer when adopting a addressed. are starting to adopt mobile devices as Building an enterprise mobility solution mobile device. productivity tools. becomes a delicate balance among security THE MOBILITY requirements, technology capabilities, device For instance, there is a significant difference CHALLENGE In the context of DSTA, Project mHabitat usability and user expectations. in the scale of deployment between an was initiated to enable access of work enterprise and a consumer. A consumer may The challenges and limitations of mobile emails on a non-camera mobile device. It The PMT’s objective is to develop a mobile pay more for faster data transmission readily. technology need to be appreciated as they started off as a trial but has evolved into a device that can provide an electronic For an enterprise, however, this additional play a part in shaping an enterprise mobility complex project. To deliver this capability, it workplace that is both secure and reliable. cost has to be multiplied by the total project such as mHabitat. was fundamental to manage and secure the number of users within the organisation. The mobile device. The need for a non-camera significant increase in cost has to be balanced THE ENTERPRISE Security Vulnerabilities and mobile device meant that the project team LANDSCAPE with the resultant increase in risk exposure either had to choose from a limited range and return on investment. Threats of mobile phones available in the market, In order to build an effective enterprise or modify an existing model. A development The main priorities of an enterprise and a The average consumer may not realise that mobility solution, it is important to first and support team was formed to establish consumer also differ. Security applications complex software and capabilities expose understand the defining attributes of an the mHabitat infrastructure. would be mandatory and of foremost priority the mobile device to a greater risk of enterprise. in the deployment of mHabitat, whereas infection from malicious software. It is often The mobile industry remains focused on design and availability of applications are a misconception that the mobile device is An enterprise is typically characterised the consumer market, creating a lag for key considerations for a typical consumer. In personal and physical possession of the device by a large user population. This leads enterprise users and an even greater lag the event where the mobile device is lost, a is required to retrieve data. In reality, the data to the challenges of cost, scalability and for security-conscious enterprises like consumer may merely feel inconvenienced can be accessed by someone else through the manageability issues as well as security of DSTA. Critical factors such as security due to the loss of personal information. following common methods: the enterprise’s data. threats, supportability, manageability and However, an enterprise usually suffers a maintainability limitations are often not much greater impact when confidential Short Message Services These concerns are different from those of a considered. Unlike laptops, mobile phones information is leaked. consumer where innovation, sleekness and rarely come with well-integrated commercial Text messages are transmitted from the other lifestyle considerations play a more off-the-shelf security solutions. Security risks From an infrastructure and support sender’s mobile device to the receiver using tangible role in the decision making process. also have to be addressed quickly but there perspective, the solution deployed has store-and-forward technology. The message are limited options in the market. to be robust and scalable to address the goes through a Cell Tower i.e. a site where 63 diverse requirements and issues brought antennas and electronics equipment are about by a large user population. Hence, placed on a radio mast or tower to create a cell developing an enterprise system would in a cellular network. It is then stored in the involve processes that are often lengthy and Short Message Service Centre (SMSC), which is Secure Mobility for the Enterprise

64

the connectivity radius of that tower. If the mobile device is able to communicate with three or more Cell Towers, a more precise location can be triangulated by comparing the signal strength and range between the towers Figure 2. Deleting a pointer to the memory address does not remove the and device as illustrated data in the physical memory in Figure 1. This information on the location benefits users a deletion process usually removes only the In an enterprise environment, a number of responsible for routing text messages before who are planning travel routes. However, if data pointers and not the actual data, unless suitable mobile devices are usually shortlisted they are forwarded. Messages that are not the same information is obtained by a third a specialised and time-consuming deletion and subjected to a lengthy and rigorous encrypted can be compromised easily during party, it can be manipulated for malicious algorithm is used. This is illustrated in Figure regression testing cycle, before they are put transit or when they are retained in the SMSC. purposes such as finding out the location of 2. The actual data that remains in the device into operation for two to three years. This is Text messages may unintentionally reveal secret installations. can be recovered easily by piecing back the no longer possible as mobile devices have a incriminating information to a third party. data structure. Software tools for recovering shorter shelf life now, resulting in a drastically Malicious software, also known as malware, Malicious software deleted information are available readily on shortened time frame available to prepare can exploit these security weaknesses and the Internet. Loss of information involving a mobile device and deploy its supporting send messages from the user’s account to A mobile device bought off-the-shelf data such as photographs and emails, usually infrastructure. premium-rate numbers which are service usually comes without any firewall or anti- occurs when the device is sent for repair or lines that charge higher prices than normal virus software. It is typically left exposed to resale. Users are often unaware of when and telephone lines. the Internet by default. Software such as Effectiveness of Technical spyware and viruses can worm their way into how the data leakage happened until it is too Support Framework late. According to a Straits Times report published these unprotected mobile devices through on 10 November 2010, the ‘Zombie’ virus that system vulnerabilities or applications installed Computers can use mainstream operating The leakage of information and other hit China affected approximately one million by the user. The risks of a trojanised mobile systems such as Microsoft Windows 7 vulnerabilities happen through various mobile phone users. The virus proliferated device are varied, ranging from unexpected and Linux. On the other hand, operating channels, ranging from a simple text automatically by sending text messages device behaviour to data loss and corruption. systems available for mobile devices are message, remote activation of cameras repeatedly to the users’ contacts. The text Beyond these risks, some malware are also customised at various stages. Thus, besides and microphones, to Trojans that have contained a link that infected the receiving known to have the ability to turn on/off the operating system developer, the mobile penetrated the mobile devices through the device upon access. The SIM card details hardware devices such as the microphone device manufacturer and telecommunication Internet. were also transmitted back to the hackers’ or the camera of a mobile device. It can service providers are also involved in the central server. The virus was estimated to cost also record a user’s conversation or take end product. This could lead to a gap in infected users S$386,000 per day. Ironically, photographs without the user’s knowledge. Mobile Device Variety and the technical support as the manufacturer’s the virus was hidden in an anti-virus The security risks are significant, especially Short Shelf Life warranty usually covers only the routine application. in confidential discussions and planning. hardware servicing. A key priority for mobile device Location Awareness In a report on CNET.com, a security agency manufacturers is to enhance the usability In February 2010, Microsoft announced fabricated user credentials and demonstrated and cosmetic appeal of their devices to that handsets which were running the Typically, the Global Positioning System how hidden applications in expansion packs attract consumers. Based on perceived market latest Windows Mobile 6.1/6.5 operating (GPS) is used to identify a person’s location. can be installed easily without the user’s trends, mobile devices are launched at an system suffered serious usability issues (e.g. However, through a mobile device’s constant knowledge (Mills, 2010). increasing rate, leading to a wide variety of the inability to answer phone calls) and communication with Cell Towers to update models that may have a short shelf life. periodic performance lags when there was 65 and maintain connectivity with the network, Data Retrieval and Forensics background traffic (Microsoft Knowledge it is possible to derive the location of a user While these factors benefit the consumer, Base, 2010). The issue took almost six months as well. A mobile device within the range of Data deleted from a computing device is they pose significant challenges and require to resolve as the responsibility was shared a single Cell Tower is assumed to be within often perceived to be irrecoverable. However, radical mindset changes from enterprises. among numerous parties who each worked Secure Mobility for the Enterprise

66

and maintain connection. Mobile device to implement. However, these measures security may be compromised and the Mobile are essential as they perform the following Device Management Software Agent that functions: relies on connectivity to send signals may be disrupted. A remote mobile device erasure Prevent unauthorised access to the feature will not work without connectivity mobile device as the device cannot receive the erase command. The solution takes the form of a Some form of password or device pin is regular data exchange between the mobile required to authenticate users in order device and the Mobile Device Management to gain access to the device’s contents. servers known as a ‘heartbeat’. The erase Two-Factor Authentication is a critical Figure 3. Abstract from Microsoft Support describing serious performance issues found in command is activated locally when a requirement for access – it may come in the Windows Mobile devices predefined number of beats are missed. form of a One-Time-Password, Password However, this option also creates another Generating Token or a Smartcard. on a specific aspect of the mobile device. like their computers. However, despite problem of distinguishing between a lost In Figure 3, an abstract from the Microsoft technology advances, the inherent form mobile device and lost network connectivity. Prevent unauthorised extraction of Support portal describes the issues faced by factor and battery capacity of a mobile device the mobile device’s data the operating system provider and hardware limit its functionality. Applications originally BUILDING ENTERPRISE manufacturer. designed for mainstream computers may Encryption of on-board and removable also not run properly on a mobile device MOBILITY storage media should be implemented To a consumer, the gap in technical support due to the difference in form factor and to protect the data from unauthorised may cause inconvenience or frustration. To an operating system. It is a challenge to design While mobile device security and extraction such as the removal of the storage enterprise, operations and productivity may an enterprise mobile device that is able to manageability are being enhanced media, and subsequent access to the contents be affected when applications fail to work or run enterprise applications, perform security continually, there is a need to manage through a media reader. However, encrypted important phone calls are missed. and device management functions as well as these fundamental gaps proactively when information is not entirely secure and can be cater to user preferences and requirements. designing an enterprise solution. For Project decrypted given sufficient time. Thus, storage Availability and Readiness of Figure 4 shows an example of the number of mHabitat, the following concepts and of sensitive data on the mobile device is not Technology and Application features and applications being processed in enablers remain important elements that recommended. a typical mobile device. define the successful implementation of an enterprise mobility solution. BlackBerry provides an end-to-end solution Prevent data leakage resulting from comprising specialised mobile phones and a Environmental Dependencies lost mobile devices device management system. While an end-to- Security and Authentication end solution facilitates deployment of mobile Mobility relies heavily on the availability Technologies such as Thin Client should devices and enhances system operability, of a reliable and high speed mobile Information security is the key priority of be implemented to minimise storage of the BlackBerry solution is proprietary and telecommunication network. In today’s the mHabitat solution. The strategies used data on the mobile device. In a Thin Client requires all information from its mobile interconnected landscape, it is not hard to against threats and attacks will dictate the environment, data are stored and processed devices to route via BlackBerry’s processing find ‘blind spots’ in fringe areas where reliability of the solution. Figure 5 shows on a remote server. For example, videos, centres in the US and Canada. This has given coverage is weak. In these fringe areas, the the concepts and actions which can be images and animations are now downloaded rise to concerns on data privacy. usability of enterprise services such as the established to protect against security threats completely before access but they are Virtual Private Network (VPN) is affected, and vulnerabilities. streamed to the mobile device. Any residual Users also expect their mobile devices to and the mobile device’s battery life is data after access can be removed from the be able to run applications and perform reduced as the device struggles to establish Security measures often compromise on the cache after a Thin Client session is closed. usability of the phone and can be difficult Concurrently, secondary measures such as

Remove Deter SECURITY THREATS Repair Prevent AND VULNERABILITY 67 Recover Protect

Figure 4. An example of a modern mobile device and its features Figure 5. Strategies against security threats and attacks Secure Mobility for the Enterprise

68

Compliance and Policy gaps for remedy. These factors ensure that Management the overall enterprise threat and risk model is not compromised. Typically, users expect a certain degree of freedom to customise their mobile devices. Software and Asset However, allowing user customisation also Management broadens the range of scenarios that can compromise security. Users may download a With many mobile devices in production, it Trojan-infected application unintentionally, will be labour intensive for IT support staff leading to data leakage. to install software or apply patches for each device manually. Users will also experience Users may also tamper with the built-in inconvenience as they need to hand over security features of the mobile device or their mobile devices to the support staff. In end up locking themselves out. As the addition, it will be an administrative burden mobile device moves away from its original to keep track of asset information such Figure 6. Defence against attacks and security risks on a mobile device configuration, the ability to prevent and as inventory levels, mobile device status, take proactive steps to resolve security software configurations and versioning. breaches is weakened. Thus, an efficient Mobile Device Management local and remote mobile device wipes should Prevent traffic sniffing and spoofing Therefore, a robust policy enforcement and a system is required to perform automated also be put in place to destroy any residual compliance audit system need to be in place or self-help software installation and data stored on the device. In addition to the VPN solution which already to ensure that enterprise mobile devices patching, as well as asset management. This provides a layer of end-to-end encryption operate within established parameters, management system optimises IT support Prevent viruses and malware between the mobile device and the enterprise even after customisation (see Figure 7). Such efficiency and facilitates the proliferation of network, more layers of data encryption (e.g. a governing system also helps to ensure deployed mobile devices. Users can choose Suitable anti-virus and firewall solutions Secure Sockets Layer) can be used between compliance with policies and maintain the to perform the tasks at their convenience or need to be installed and updated frequently. the device and the application to prevent consistency of permitted mobile device allow these tasks to be automated without Physical components which are not required traffic sniffing and spoofing attempts. configurations for all users, as it facilitates any user intervention. (e.g. camera, GPS) should also be disabled quick identification of non-compliance and permanently to protect the mobile device Preventive measures such as prohibiting from malware that can enable tracking and camera devices within military establishments eavesdropping. The operating system should play a vital role in maintaining information also be hardened. This is done by eliminating security. Other features such as voice and text unnecessary programs to reduce the security encryption as well as microphone treatment vulnerabilities. offer additional layers of security at the software and hardware levels (see Figure Prevent zero-day attacks on internal 6). However, the implementation of these resources from the Internet security features is dependent on the nature of the mobile device. For example, if the Access to all resources should be via a VPN source codes and Application Programming tunnel so that the resources do not have Interface of the mobile device’s operating direct access to the Internet. The single point system are unavailable, programmers may not of entry also helps to ensure a consistent have the ability to modify the device for voice security posture, which facilitates ease of and text encryption or exchange of data. monitoring network traffic that enters the device and restricts entry to malicious websites. 69

Figure 7. An example of mobile device policy management Secure Mobility for the Enterprise

70

The system can also improve security by unwanted attention or embarrassment. regulating and monitoring applications The crushing or smelting of a mobile device Internal Factors External Factors which are installed on the mobile device. guarantees total destruction of any residual User Requirements Market forces Thus, the installation of unauthorised or data. Enterprise Business Requirements malicious software can be prevented. Mobility Security Threats Cost and Budget Infrastructure and Application Solution Technological In addition, the system should provide a Interoperability Corporate Policies Advancement history of software installation and patching. Centralised monitoring of resource usage and An enterprise mobility solution seldom Figure 8. Factors which may shape the final mobility solution trend patterns can facilitate licence tracking, works in isolation. It usually needs to audit processes and technical support. interface with the existing infrastructure of an enterprise to provide applications and Mobile Device Support Model services to the users. With a wide variety CONCLUSION REFERENCES of mobile devices and a growing number Users typically expect to have access to of mobile enterprise platforms Enterprise mobility is a challenging Microsoft Knowledge Base. Some GUI their enterprise mobile devices regardless available, it is crucial to identify the issue. Factors such as user expectations, Interactions No Longer Work Correctly on of location and time. Correspondingly, they solution that can best fit into the technological maturity and product Windows Mobile 6.1 Devices and Windows expect technical support to be available environment. This maximises compatibility availability often determine the Mobile 6.5 Devices. http://support.microsoft. constantly. This is a paradigm shift from the while reducing unnecessary administrative solution that will be deployed (see com/kb/979220 (last modified 12 February conventional technical support of personal overheads and system conflicts. Usability Figure 8). However, these factors are 2010) computers (PC) and laptops, where users would also be greatly enhanced if the often incomplete, contradictory and bring their laptops to the IT support centres application is designed properly to run on inconsistent. Due to these complex Milanesi, C., Jones, N., Basso, M., Gammage, or have the IT support staff make a trip to the mobile device. interdependencies, the effort to solve one B., De La Vergne, H.J., Hart, T.J., Miguel, their workstations to troubleshoot PC issues. aspect of the problem may lead to other E.S., Nguyen T.H., Gupta A., Dulaney, K., Technology Limitations problems. Zimmermann A., Tully J., Ohr S., Cozza R., If technical issues are not resolved in time, Shen S., Teng A., Erensen J. and Leong A. 2009. users who have limited tolerance for device An effective enterprise mobility solution is Hype Cycle for Mobile Device Technologies. Mobile technology evolves at a rapid pace downtime are likely to turn to self-help one that bridges these gaps while allowing Gartner Inc. http://www.gartner.com/ and renders today’s solution outdated solutions which might create a configuration room for growth and flexibility. In addition, DisplayDocument?id=1085512 (accessed 23 tomorrow. The mobile device today may drift and other downstream issues. it is able to run on different mobile devices July 2010) not have enough processing and storage and operating systems while addressing capability to perform all the management Hence, to cater to people who are constantly fundamental usability, management, security Mills, E. 2010. Google Pulls App that Revealed and security tasks required. Similarly, the on the move with their mobile devices, and support requirements. An infrastructure Android Flaw, Issues Fix. CNET News, underlying telecommunication network may the IT support role needs to be expanded. that supports a single mobile device and 11 November. http://news.cnet.com/8301- not be robust enough to support the data Effective and efficient technical support can operating system is no longer sufficient. 27080_3-20022545-245.htm (accessed 12 transfer involved. However, the same cannot be achieved through mechanisms such as February 2011) be said for tomorrow. remote device management. The ultimate objective of Project mHabitat is to implement an effective enterprise Apart from piecing together the various mobility solution that is a productivity Device Life Cycle Management management and security components multiplier for the organisation. With a required, it is important to ensure that lightweight, seamless and mobile means of Mobile devices have very short shelf lives and these components minimise the computing accessing business applications, DSTA users the task of managing replacements is often overheads so that the mobile device can would be able to reap the tangible benefits overlooked. An efficient process aligned continue to do what it was really intended of true enterprise mobility. with the fast-paced mobile industry needs for. At the same time, the eventual to be adopted in order to facilitate prompt solution must be competent enough to selection and operationalisation of withstand telecommunication problems replacement mobile devices. Proper such as intermittent connectivity or poor management and disposal of the obsolete performance. 71 mobile devices are also needed to prevent Secure Mobility for the Enterprise

72

BIOGRAPHY

Adrian Toh Thiam Huat is a Principal Engineer (Infocomm Infrastructure) and he manages the Infocomm Infra Messaging Programme. He oversees the development, implementation and maintenance of the Corporate IT Messaging Systems for the Ministry of Defence (MINDEF) and DSTA, as well as the Enterprise Server Farm for DSTA. His OA Email Disaster Recovery project which enables the restoration of a messaging system within a stipulated time clinched the MINDEF Corporate IT Award in 2009. Adrian has obtained certification in Project Management and IT Business Continuity Management and is also a certified Data Centre Professional. He is currently on the Board of Assessors for Singapore Computer Society Certification in the IT Business Continuity Management Programme. Adrian obtained his Bachelor of Science (Business IT) degree with Honours from University of Central England, UK in 1998.

Tan Shyh Hae is a Senior Engineer (Infocomm Infrastructure). He is currently pursuing a Master of Science (Computing and Security) degree at King’s College, UK under the DSTA Postgraduate Scholarship. He led a team overseeing the requirements development and architecting of messaging projects, specialising in mobile related solutions. Shyh Hae obtained his Bachelor of Engineering (Computer Engineering) degree from the National University of Singapore in 2000.

Lim Dawei is an Engineer (Infocomm Infrastructure). He is involved in the deployment of DSTA Internet Access Infrastructure and is researching and deploying Virtual Desktop and Thin Client Technology to achieve greater enterprise mobility with enhanced security. Dawei also assesses and mitigates risks of mobile computing. He was involved in the deployment of messaging systems for MINDEF. Dawei obtained his Bachelor of Engineering (Computer Science) degree with Honours from Nanyang Technological University in 2008.

73 ABSTRACT Capability Development Strategic management literature has recognised the instrumental roles of the market’s external requirements and Framework for the firm’s internal capabilities in driving innovation. Defence Technology Investments Capability development has been identified as a key objective of defence technology investments. This article presents a framework that maps this capability development process to real options. The real options embedded in defence technology investments evolve as uncertainty decreases with technological maturity, and as the readiness for field transitions increases with clearer technological applications. Historical examples of important defence technological innovation and contemporary examples of technological innovation by DSTA are cited to illustrate the framework. The article concludes with a discussion on framing the capabilities as real options.

Ang Choon Keat Lee Keen Sing Ng Si Jing Capability Development Framework for 76 Defence Technology Investments

to meet expected market applications Technology Readiness Level (TRL) INTRODUCTION in the foreseeable future, technology investments also serve to develop firm- 1 Basic principles observed and reported There has been extensive research conducted specific capabilities (Helfat, 1994) and the 2 Technology concept and/or application formulated to understand the dynamics of technological means to sustain competitive advantage 3 Analytical and experimental critical function and/or characteristic proof of concept development. Strategic management for unpredictable long-term requirements literature suggests that innovation can be (Clarke and Pitt, 1996; Cohen and Levinthal, 4 Component and/or breadboard validation in laboratory environment driven by external requirements of the 1989; Kogut and Kulatilaka, 2001). For 5 Component and/or breadboard validation in relevant environment market (Schmookler, 1966), and by the example, technological capability reflects a System/subsystem model or prototype demonstration in a relevant environment (ground or activities and internal capabilities of firms firm’s strength in discovery and innovation, 6 space) (Dosi, 1982). Hence, besides technological and enables it to value, assimilate and exploit development, the creation of capabilities is new knowledge (Cohen and Levinthal, 1989). 7 System prototype demonstration in a space environment crucial in a technology investment (Clarke and The evolution of capabilities can be modelled Actual system completed and ‘flight qualified’ through test and demonstration (ground or 8 Pitt, 1996; Helfat, 1994; Cohen and Levinthal, by a life cycle involving the stages of space) 1989). In defence technology investments, founding, development and maturity 9 Actual system ‘flight proven’ through successful mission operations the development of indigenous defence (Helfat and Peteraf, 2003). In particular, a technological capabilities is a strategic strategic objective of defence technology Table 1. TRL in NASA objective (Ang and Chai, 2009). This article investments is the development of builds on these discussions and presents indigenous defence technological technological capabilities. They offer the Transformation Map a capability development framework for capabilities (Ang and Chai, 2007), while the end user technological options i.e. the defence technology investments. Specific R&D process can be framed as a capability right but not the obligation to develop In defence technology investments, the capabilities driven by application can be development process (Ang and Chai, 2009). the technological capability into a system large-scale mission-oriented projects aim framed as technological options (Mitchell capability. These capabilities would be of to develop specific technologies under and Hamilton, 1988), while more broad- Capability Transformation TRL 5 and 6. These technological options conditions of high appropriability and high based options can be framed as a generic set correspond to the real options in strategic cumulativeness at the firm level (Malerba, of resources creating platforms for future Process positioning proposed by Mitchell and 2004). These lead to a Schumpeter Mark developments and opportunities (Kogut and Hamilton (1988). An example is the R&D can be viewed as a process of resource II Model of innovation regime (Breschi, Kulatilaka, 1994; 2001). exploratory development in specific transformation (Schmidt and Freeland, Malerba and Orsenigo, 2000) characterised technologies. 1992) where firms create strategic options by “creative accumulation” and the LITERATURE REVIEW by transforming resources into capabilities importance of experience gained from Knowledge of the Firm. This refers to the which offer strategic flexibility. This innovation efforts. Noting that the R&D compound options created from investment Capability Development from capability development process can be process involves the development of in knowledge of the firm. An example is Technology Investments modelled as a capability life cycle (Helfat and capabilities which could be modelled as investment in human capital. The knowledge Peteraf, 2003). Ang and Chai (2009) propose a capability life cycle (Helfat and Peteraf, created can be considered as owning a 2003), Ang and Chai (2009) proposed a The multiple benefits from technology that the capabilities developed in defence portfolio of options or platforms for future transformation map for this capability investments have been well discussed R&D evolve and can be categorised into developments (Kogut and Zander, 1992). This development in defence technology in strategic management literature. technology maturity levels as follows: would correspond to TRL 4 and below. investments (see Figure 1). Practitioners such as Andrew and Sirkin (2007) recognise the indirect benefits of Developmental Capabilities. These innovation such as knowledge acquisition. capabilities are required to develop Low application uncertainty High application uncertainty Knowledge and innovation are cumulative technology applications for operational and evolutionary (Nelson and Winter, 1982). requirements such as enhancing operational High Application- Thus, technology investments enable firms capabilities to improve a weapon system. technological driven capability development to come up with incremental innovations, Using the Technology Readiness Level (TRL) uncertainty which eventually culminate in the creation framework developed by the National Spiral Development Aeronautics and Space Administration of technological variation or the quick development of technological adoption of technological changes. This (NASA), developmental capabilities would Low options allows the firm to move in tandem with the correspond to TRL 7 and above. Table 1 technological 77 unpredictable technological discontinuities illustrates this. uncertainty which punctuate the technological life cycle of any system (Tushman and Anderson, 1986). Technological Capabilities. These are Besides developing particular technologies vanilla options created from investment in Figure 1. Transformation map of capabilities Capability Development Framework for Defence Technology Investments 78

The transformation could be seen as had to drive a drill bit into the bottom of the a development vector describing the hull of the target vessel, attach a waterproof Low application uncertainty High application uncertainty maturation of the technology and also time bomb and then escape before the bomb the resolution of uncertainty in the was detonated by a clockwork fuse. Driven by military application, the Traditional naval warfare application. Different driving forces behind High capability of the submarine improved occurs on the water the capability development would lead to The Nautilus was designed by Robert Fulton. technological from the first workable submarine, surface. the Turtle. An improved submarine, the development vector taking different During successful demonstrations in 1801 uncertainty H.L. Hunley, sank a target in battle paths. Thus, the real option embedded in a and 1805, it was able to cruise under the successfully. technology development programme would intended victim and tow the explosive bomb evolve accordingly. Within this framework, until the bomb came in contact with the one can examine the relationship among target and detonated with a contact fuse. The first practical combat submarine Low was produced, introducing defence technology investments, capability This craft had a copper-sheathed hull and technological innovations such as the diesel engines, development and options creation for the it was equipped with a mast, bowsprit and uncertainty improved periscopes and torpedoes, uncertain future. The framing of defence two sails for surface propulsion and two and wireless technology. technology investments as real options in hand-cranked screws to travel underwater capability development underscores the (see Figure 2). Depth was estimated using Figure 3. Transformation map: Application-driven capability development of submarines theoretical foundation for the application a barometer, while air was supplied to the of real option theory to model defence four-man crew by flasks of compressed air on Rockets combined with a fuel that had a very high technology investments. board. and powerful burn rate, such as hydrogen. The Chinese were early users of gunpowder The subsequent sections of the article and invented gunpowder-propelled rockets Goddard’s work inspired a group of German illustrate this framework with historical in the early 13th century. Many subsequent rocket enthusiasts to adopt his technical ideas examples of important defence technological military thinkers and technicians dreamt of for their own rocket experiments. In 1935, innovation and contemporary examples of giant rockets that could be launched to hit this group of rocket enthusiasts was enlisted technological innovation by DSTA. targets hundreds of miles away. However, by the German army to develop long-range the gunpowder propulsion was insufficient ballistic rockets capable of carrying large CASE STUDIES: to propel a heavy rocket over any significant explosive warheads. During World War Two HISTORICAL EXAMPLES distance. In addition, the rocket could not be (WWII), the group developed the V-2 rockets launched beyond the earth’s atmosphere as which produced 28,000kg of thrust. A fuel of liquid oxygen and alcohol together with Data Selection and Analysis the gunpowder would have no oxygen to Figure 2. Artist’s impression of Nautilus towing burn for its propulsion. a set of gyroscopes and flight guidance explosive bomb (Source: War and Game) fins could launch a 400lb warhead of high The data selected are the capability Robert Goddard demonstrated in 1919 that explosives towards a target hundreds of miles development processes leading to some of In 1900, John Holland won a submarine these problems could be overcome by a rocket away. Figure 4 shows the transformation the most important defence technological design competition held by the US Navy carrying its own oxygen supply in liquid form, map of the application-driven capability innovations (van Creveld, 1991; Perry, 2004). and went on to design the USS Holland development of rockets. The capability development process for each (SS-1), the first practical combat submarine. innovation is analysed and mapped onto the It included innovative features such as self- transformation framework. propelled torpedoes fired from a reloadable Low application uncertainty High application uncertainty tube, a battery-powered electric motor for Submarines submerged operations, and an advanced hull The Chinese invented rockets propelled High shape to allow it to move efficiently through by gunpowder. However, giant rockets Invention technological Naval warfare has traditionally been waged the seas. Several innovations in military could not be launched to hit targets of gunpowder a significant distance away due to through caravels, galleons, men-of-war uncertainty submarines were made in the period before insufficient gunpowder propulsion. and frigates on the water surface. The World War One (WWI) – these included the capability development for a submarine to development of diesel engines, improvement attack a surface vessel from underwater was in periscopes and torpedoes, as well as the Low Goddard demonstrated rocket propulsion driven primarily by military application. The advancement of wireless technology which using liquid fuel. His technical ideas were 79 first workable submarine, the Turtle, was enabled submarines to be directed from shore technological further developed by the Germans who designed by David Bushnell in 1776. It was bases. The transformation map in Figure 3 uncertainty went on to build the V-2 rockets during propelled by a hand-crafted screw and had illustrates the application-driven capability WWII. room for only one crewman. This crewman development of submarines. Figure 4. Transformation map: Application-driven capability development of rockets Capability Development Framework for Defence Technology Investments 80

Low application uncertainty High application uncertainty Low application uncertainty High application uncertainty

High The reflection of radio High Invention of internal waves from a metallic object technological combustion engine and technological was demonstrated. uncertainty caterpillar track uncertainty

Watson-Watt demonstrated The use of armoured assault that an aircraft could be detected by radar waves The essentials of the practical Low vehicles was explored. The first Low tanks were built in Britain using The internal combustion engine with its position, altitude and radar were complete with technological the engine and the transmission and caterpillar track were used technological course plotted on the CRT. The the discovery, generation and uncertainty of wheeled tractors, as well as in early farm tractors. uncertainty British set up the Chain Home detection of radio waves as well the tracks of tractors. Stations to scan the eastern as the commercial availability of and southern skies. the CRT.

Figure 5. Transformation map: Development of technical options for tanks Figure 6. Transformation map: Development of technical options for radar (Note: Radar does not generally depend on reflection from the ionsphere, but the ionsphere was the first object detected by radar)

By 1938, the British Chain Home Stations set Many lessons were learnt – observers were first ever tank action to help the infantry Tanks up to scan the eastern and southern skies needed to take notes of ground activity, assault enemy trenches on the Somme. The were reaching out with 60% reliability to 70 and more pilots and aircraft had to be transformation map in Figure 5 illustrates the The key enabling technologies for the miles at 20,000ft. A chain of radar stations was available. These in turn required a better development of technological options that tank – internal combustion engine and built along the South and East coasts of Britain servicing organisation. Better maps were also led to the military innovation of tanks. caterpillar track – were mature technologies by 1939. Linked to a highly efficient control needed, leading to the development of aerial used in early farm tractors before the network, this early radar system played a photography. Thus, the Libyan campaign military innovation of tanks during WWI. Radar crucial part in detecting formations of enemy taught the Italians the usefulness, rapidity During the war, opposing armies reached a aircraft as they approached the coast. As the and reliability of air reconnaissance, as well deadlock as the traditional infantry attacks In 1934, Robert Watson-Watt of the National Fighter Command was allowed to deploy its as the need for accuracy in bombing, the had become challenging due to increasingly Physical Laboratory informed the British Air resources most effectively, the success of the dangers of ground fire, and the limitations of effective firepower, as well as the extensive Ministry that an aircraft could be detected Battle of Britain was secured. Figure 6 shows equipment. deployment of entrenchment and barbed at long range by radar waves. On the the development of technological options wire for defence. Consequently, the use of cathode ray tube (CRT) screen which had that led to the innovation of radar systems. With the deadlock experienced during WWI, armoured assault vehicles was explored as been commercially available since 1922, the reconnaissance aircraft were the only means they could crush the barbed wires and protect aircraft could be displayed showing three Military Aircraft of gaining information on the position against machine gun fire while approaching key parameters: its position (coordinates), of enemy artillery and reserves. Fighter enemy trenches. altitude and course plotted. After the Wright brothers demonstrated the aircraft naturally evolved as a means first heavier-than-air powered flying machine of denying the enemy this invaluable The first experimental tank was built in The reflection of radio waves from a metallic controlled by an on-board pilot on 17 information by arming aircraft to bring down Britain in September 1915 using the engine object was first demonstrated in 1855 and December 1903, the more advanced military other planes. However, early gunnery was and the transmission of wheeled tractors as the ionosphere discovered in the early powers including the US and UK were not primitive and the pilots were armed only with well as the tracks of Bullock tractors procured 1920s provided the essential pre-requisites keen to develop aircraft for the next pistols and hand grenades. A suitable aerial from the US. An improved design with a for the development of radar. Using the three years. Nonetheless, the enthusiasts weapon would be a forward-firing machine much longer track was developed to enable principle that solid objects reflect radio experimented with dropping bombs, gun, but the bullets would hit the propeller the tanks to cross obstacles such as trenches waves, sending radio waves out on a fixed installing machine guns as well as mounting blades as the gun was sited parallel to the which were up to 1.5m in width and parapets wavelength and recording the echo made it aerial photography equipment, and aircraft fuselage. which were up to 1.4m in height. This was possible to calculate the range and direction demonstrated many modern functions of 81 completed and demonstrated successfully in of the object’s movements. In February 1935, air power. Although the aircraft only had Early experiments tried to overcome this February 1916, and the War Office ordered Watson-Watt demonstrated the detection of primitive capabilities, the Italians used them problem by fitting deflectors onto the 150 similar vehicles. On 15 September 1916, an aircraft flying at 10,000ft over a range of in the 1911 war against the Turks in Libya to propeller blades to deflect any bullets that the 49 tanks available were sent on the 13km. observe ground activity. hit the propeller blades away from the pilot. Capability Development Framework for Defence Technology Investments 82

attendant oil products. The transformation CASE STUDIES: map in Figure 9 shows the application-driven CONTEMPORARY capability development of the UAF. EXAMPLES Infrared Fever Scanner System Underground Ammunition Facility In response to the Severe Acute Respiratory Syndrome (SARS) crisis in 2003, there was an urgent requirement by the health The limited land resources for ammunition authorities to identify individuals with body storage facilities resulted in the construction temperatures that were higher than average. of the Underground Ammunition Facility The usual method of using thermometers to (UAF) for the Singapore Armed Forces Figure 7. Deflector on propeller blades and a machine gun fitted with an interrupter gear measure temperatures was tedious and time- (Source: Fiddlers Green) (SAF). There was no precedent of a large- consuming. This led to the opportunistic use scale underground ammunition facility of the mature infrared technology for the Although successful, this method impaired against German Zeppelin sheds near Metz. developed within a densely populated and Infrared Fever Scanner System (IFSS) as shown aiming and many shots were wasted as the Typical of the early bombers, the Voisin urbanised area. Thus, significant technology in Figure 10. machine gun often fired into the deflector was basically a general purpose aircraft investments were required to meet rather than through the propeller blades. from which up to 124lb of bombs could be the protective infrastructure requirements The Germans solved the problem eventually dropped by hand. Hence, the development and develop the related technologies, with a proper interrupter gear that enabled priorities for bomber aircraft were greater which could be harvested as feasible options the pilot to fire fixed guns at random power and speed to improve on range and to ensure explosion containment. The through the propeller arc (see Figure 7). This payload, as well as accurate navigation and knowledge gained from the successful mechanism was incorporated in the Fokker bombsights. Midway through the war, the technology development and application Eindecker 1 Fighter Aircraft by the summer of Italians developed the large Caproni Ca series in the construction of the underground 1915 and effectively tilted the air warfare in which had a ceiling of 13,400ft in its later cavern paved the way for the construction favour of Germany, until the Allies’ aircraft versions. It was also capable of speeds up of underground caverns of similar scales. were similarly equipped with an effective to 85mph and bomb loads of up to 1,000lb. This knowledge was used to set new safety interrupter gear in mid-1916. The Italians became the first to carry out true standards and contributed to the North strategic bombing, amassing large numbers Atlantic Treaty Organisation safety codes. In An equally significant development was the of aircraft to strike against a single target. addition, the suite of technologies developed advent of the bomber aircraft and the surge The spiral development of bomber aircraft and applied for the UAF could possibly in demand for its bombing function. The first is illustrated in the transformation map in be adapted for other uses, such as using Figure 10. IFSS bombing raid of the war was carried out by Figure 8. underground caverns to store crude oil and French Voisin bombers on 14 August 1914

Low application uncertainty High application uncertainty Low application uncertainty High application uncertainty

No precedent of a large-scale Spiral 2: Development of fighter High Significant R&D investments Spiral 1: Enthusiasts underground ammunition facility High aircraft was initiated as a means in protective infrastructure experimented and technological developed within a densely to bring down other aircraft. and related technologies technological demonstrated many populated and urbanised area Bomber aircraft development uncertainty modern functions of air uncertainty was initiated and the concept power. The training of of strategic bombing advanced. pilots and manufacturing of aircraft were primitive.

Spiral 1: The Italians took the Low Capability for the construction Low Spiral 2: The technical problem primitive aircraft to war in 1911 of mounting machine guns technological of underground caverns for technological with the main task of observing ammunition storage on fighter aircraft was solved ground activity, and they learnt uncertainty 83 uncertainty with the interrupter device. important lessons. Specialised bomber aircraft were produced. Figure 9. Transformation map: Application-driven capability development of the UAF Figure 8. Transformation map: Spiral development for bomber aircraft Capability Development Framework for Defence Technology Investments 84

Low application uncertainty High application uncertainty Low application uncertainty High application uncertainty

Spiral 2: Continued R&D into Spiral 1: R&D into UAV High Infrared technology High UAV development for larger technology was initiated was a mature technological class of tactical UAV called in DSO about a decade technological technology. uncertainty Skyblade IV ago to build up indigenous uncertainty capability in unmanned aircraft technology.

Spiral 1: Skyblade III Mini-UAV Low During the SARS crisis, a Low Development of the IFSS means to identify subjects was developed successfully from technological technological R&D to operationalisation for with temperatures higher than Spiral 1: Extensive field trials and the Army uncertainty normal was explored. uncertainty design evolution were undertaken to overcome technical challenges.

Figure 11. Transformation map: Development of technical options for IFSS Figure 12. Transformation map: Spiral development of UAV

Due to the urgent requirement, the initial events (Kogut and Kulatilaka, 2001). Since Indigenous Unmanned Aerial DISCUSSION IFSS categorised the subject’s temperature Vehicle R&D investments are the technological based on shades of colour as a proxy. This equivalent of a financial options contract, has since evolved to the numeric tagging The case studies illustrate the transformation technology investments with high levels R&D efforts to build up indigenous of temperature to the subject’s forehead as map framework which examines the of uncertainty are better assessed using capability in Unmanned Aerial Vehicle (UAV) one appears on the sensor computer screen. relationships among defence technology real options (Mitchell and Hamilton, 1988). technology was initiated in the early 2000s. The technology for numeric tagging was investments, capability development and Such projects offer the investor the right Initial efforts were focused on developing a already well developed in other applications. options creation for the uncertain future. but not the obligation to create a product man-portable mini-UAV called Skyblade I to The use of this technology provided better at some point in the future, in return for a support Army battalion operations. However, resolution and accuracy as compared to the limited downside in investment. This technical challenges were encountered as Capability Development use of colours as a proxy. The IFSS proved to is substantially useful during times of the subsystems had to be small and Process be useful in the temperature filtering process uncertainty. lightweight, yet robust and reliable. Extensive to contain the spread of the H1N1 virus. field trials and design revisions were carried The capability development processes for Kogut and Kulatilaka (1994; 2001) argue out before Skyblade III progressed successfully submarines and rockets were driven by While the application of the infrared that capabilities are real options as they from an R&D prototype to a production military application. On the other hand, technology for temperature screening was form platforms that create a generic set model. The product was a joint effort the military innovations of the tank and opportune, it has not been proven that of resources and represent investments in between DSO National Laboratories (DSO) radar were the result of application- the infrared technology will be reliable future opportunities. A heuristic framing of and Singapore Technologies Aerospace, and driven development, where key enabling under different operating conditions such capabilities as real options was proposed to it became the first indigenous UAV to be technologies were developed and as high traffic flow, dust, humidity and guide the normative evaluation of deployed by the SAF in 2009. The knowledge matured independently beforehand. The ambient temperature. More R&D and trials exploitation and exploration. Noting that and experience gained from this R&D effort capability development for military aircraft are required to ensure that the technology technological options differ in strategic is now channelled into the development of a demonstrated a different process in which is sufficiently mature to be deployed for purposes and have different levels and types larger class of tactical UAV called Skyblade IV. new applications and requirements for temperature screening under different of uncertainty involved, MacMillan and The spiral development of the UAV is shown technological development were discovered operating conditions, with an acceptable false McGrath (2002) propose that R&D projects in Figure 12. through spiral experimentation and a alarm rate. Figure 11 shows the development learning process. should be treated as one of the three types of technological options that led to the of real options, depending on their degree of innovation of the IFSS. Capabilities as Real Options technical and market uncertainty: 85 a) Positioning options are taken out to Real options are investments in physical preserve a company’s opportunity to compete assets, human competence, and in future technological arenas that are still organisational capabilities that provide the unclear. opportunity to respond to future contingent Capability Development Framework for Defence Technology Investments 86

breakthroughs which would give an edge Low application uncertainty High application uncertainty over adversaries. On the other hand, defence technology investments in mature Positioning options Stepping-stone options technologies may aim to create incremental High innovations in existing applications. “Modular innovation” “Radical innovation” Nonetheless, with creative accumulation in technological e.g. quantum leap in weapon e.g. R&D investments in emerging the evolution of technology, options systems performance breakthrough technology which uncertainty embedded in a technology development would influence the outcome of the war programme evolve as (a) the technological uncertainty decreases with technological maturity, and as (b) the readiness for field Enhancement and platform Scouting options transition increases with identification of Low launches technological “Application innovation” applications. uncertainty “Incremental innovation” e.g. fielding existing technologies e.g. upgrading weapon systems in new doctrines of operation CONCLUSION

Table 2. Technological and scenario uncertainties in defence technology investments A strategic objective of defence technology (Source: Modelled after MacMillan and McGrath, 2002) investments is the development of indigenous defence technological capabilities, while the R&D process can be viewed as a capability development b) Scouting options are used to learn enabling technologies have been developed process. This article illustrates the capability about the market by probing or offering and matured independently beforehand, development framework proposed by prototypes to potential early adopters. sometimes outside a military laboratory. For Ang and Chai (2009) through historical example, the internal combustion engine examples of important defence technological c) Stepping-stone options are created and caterpillar track used in the innovation innovation and contemporary examples of when market and technological uncertainties of tanks were mature technologies used in defence technological innovation by DSTA. are high to build both market insights and early farm tractors. This capability development framework technical competence systematically. contributes to the literature on the dynamics The innovation of tanks also illustrates of technological strategy innovations. The Framing the R&D project as one of these three the importance of certainty in fielding the strategic heuristic can be examined through options will enable a company to progress application. While the caterpillar tractors the relationships among defence technology without exposure to potentially catastrophic had been used in military service, either as investments, capability development and downside risks. In highly unpredictable a means of hauling cargo or a device for options creation for the uncertain future. situations, smart companies have learnt that pulling very large artillery pieces, few people Specific capabilities driven by application can the deployment of a portfolio of options is thought of arming caterpillar tractors as be framed as technological options, while the best way to ensure their ability to respond there was no requirement before WWI. It more broad-based options can be framed as effectively to future challenges. was only after the deadlock situation at the a generic set of resources creating platforms trenches of the Western Front that armies for future developments and opportunities. Strategic Heuristic realised the importance of fielding an armoured fighting vehicle. In the particular case of defence technology investments, the portfolio would likely The strategic purposes would likely vary for include projects with different levels of defence technological projects with different uncertainty in developing the technology levels of uncertainty in developing the and fielding the application (see Table 2). technology and in fielding the application. The level of uncertainty of a greenfield Defence technology investments in technology is high. Conversely, the level of areas with high uncertainty in development 87 technological uncertainty could be low if the and application may aim to create Capability Development Framework for Defence Technology Investments 88

Kogut, B. and Kulatilaka, N. 2001. Capabilities REFERENCES as Real Options. Organization Science 12(6): BIOGRAPHY 744-758. Andrew, J.P. and Sirkin, H.L. 2007. Payback: Reaping the Rewards of Innovation. Harvard Kogut, B. and Zander, U. 1992. Knowledge Ang Choon Keat is a Senior Engineer (Building and Infrastructure) and works Business School Press. of the Firm, Combinative Capabilities and on planning, designing and managing protective building and infrastructure the Replication of Technology, Organization developments. He is a registered Professional Engineer (Civil) in Singapore. Ang, C.K. and Chai, K.H. 2007. Real-Options Science 3(3). Choon Keat was previously Manager (Plans) at the Defence Research and Based Approach to Management of Defence Technology Office where he was involved in the management of the Ministry R&D Investments: An Exploratory Study. MacMillan, I.C and McGrath, R.G. 2002. of Defence Research and Technology portfolio. He graduated with a Master Proceedings of the Asia-Pacific Systems Crafting R&D Porfolios. Research-Technology of Engineering (Civil Engineering) degree from Imperial College London, UK Engineering Conference 2007. Management 45(5):48-59. in 1999 under the DSTA Undergraduate Scholarship. He further obtained a Master of Science (Operations Research) degree from Columbia University, Ang, C.K. and Chai, K.H. 2009. A Real Options Malerba, F. 2004. Sectoral Systems of US in 2005 under the DSTA Postgraduate Scholarship. Choon Keat is currently Framework for Defence R&D Investments in Innovation: Concepts, Issues and Analyses pursuing a Doctor of Philosophy (Management of Technology) degree from Capability Development. Proceedings of the of Six Major Sectors in Europe. Cambridge the National University of Singapore. Asia-Pacific Systems Engineering Conference University Press. 2009. Mitchell, G. and Hamilton, W. 1988. Breschi, S., Malerba, F. and Orsenigo, L. 2000. Managing R&D as a Strategic Option. Lee Keen Sing is a Principal Engineer (Land Systems) and leads a team in Technological Regimes and Schumpeterian Research Technology Management 31(3): innovating technological solutions and modifying equipment in response Patterns of Innovation. Economic Journal 15-22. to requirements of the Singapore Armed Forces (SAF). He is also a Senior 110:338-410. Investment Manager of Cap Vista Pte Ltd which identifies and invests in Nelson, R.R. and Winter, S.G. 1982. An companies that develop cutting-edge technologies to support Singapore’s Clarke, K. and Pitt, M. 1996. R&D Initiatives Evolutionary Theory of Economic Change. defence initiatives. He was previously involved in the areas of Systems and the Development of Strategic Harvard University Press. Architecting, Capability Development and R&D. Keen Sing obtained a Advantages. R&D decisions: Strategy, Policy Master of Engineering (Mechanical and Production) degree from Nanyang and Innovations. Routledge. Perry, W.J. 2004. Military Technology: An Technological University (NTU) under the NTU Research Scholarship and Historical Perspective. Technology in Society the National Science and Technology Board Postgraduate Training Initiative Cohen, W.M. and Levinthal, A. 1989. 26:235-243. scheme in 1999. Under the DSTA Postgraduate Scholarship, Keen Sing Innovation and Learning: The Two Faces of graduated with a Master of Science (Engineering and Management) degree R&D. Economic Journal 99:569-596 Schmidt. R.L. and Freeland, J.R. 1992. Recent from the Massachusetts Institute of Technology, US in 2004. Progress in Modeling R&D Project-Selection Dosi, G. 1982. Technological Paradigms and Processes. IEEE Transactions on Engineering Technological Trajectories. Research Policy Management 39(2). 11:147-162. Ng Si Jing is an Analyst (DSTA Masterplanning and Systems Architecting). Schmookler, J. 1966. Invention and Economic She examines how technology options may impact the effectiveness of SAF Helfat, C.E. 1994. Firm-Specificity in Corporate Growth. Harvard University Press. R&D. Organization Science 5:173-184. missions. She was involved in the local development of unmanned aerial vehicles. A recipient of the DSTA Undergraduate Scholarship, she graduated Tushman, M.L. and Anderson, P. 1986. with a Bachelor of Science (Mechanical Engineering) degree from Cornell Helfat, C.E. and Peteraf, M.A. 2003. The Technological Discontinuities and University, US in 2004 and a Master of Science (Aeronautics and Astronautics) Dynamic Resource-based View: Capability Organizational Environments. Administrative degree from Stanford University, US in 2005. Lifecycles. Strategic Management Journal Science Quarterly 31(3):439-465. 24(10):997-1010. Van Creveld, M. 1991. Technology and War: Kogut, B. and Kulatilaka, N. 1994. Options From 2000 B.C. to the Present. Touchstone. Thinking and Platform Investments: Investing and Opportunity. California Management 89 Review 36:52-71. ABSTRACT Comparative Analysis of Radar and sonar are sensors which are used widely in the military domain. Research and development in these areas are Radar and Sonar Principles normally carried out independently by different groups. While each sensor system has its own set of challenges and solutions, there are fundamental similarities in their operating principles. This article examines and compares these similarities to provide meaningful insights into the synergy between radar and sonar.

Teo Seow Khye Neo Chin Sim Solomon Sia Hou You Comparative Analysis of Radar and Sonar Principles 92

processing and analysis of the signal received INTRODUCTION enable the system to detect the target as well as determine characteristics such as range, Radar and sonar have seen extensive military bearing, course and speed. use in their primary role of detecting and identifying varied threats. Both are sensor systems which use the transmission and TYPES OF RADAR AND reception of return signals to function. Radar SONAR systems operate using radio waves primarily in air, while sonar systems operate using The types of radar range from those installed sound waves primarily in water (Minkoff, on large static platforms with detection 1991). Despite the difference in medium, ranges over 100km, to miniaturised versions similarities in the principles of radar and installed on unmanned aerial vehicles that sonar can frequently result in technological offer situational updates in a localised area convergence. of operations. Three distinct types of radar systems are the static early warning area Although radar and sonar operate under surveillance platform, the targeting and different environments and take on fire control platform, and the battlefield various roles, they are closely related due reconnaissance detection and search to fundamental similarities in function. platform. Radar detection can be active or passive, and the latter utilises the Therefore, breakthroughs in one field Figure 1. The Doppler effect: Velocity’s impact on frequency (Source: European Space Agency) can potentially provide insights and triangulation method based on lead to the advancement of the other. This electromagnetic reception or the bistatic For short-range scanning of high resolution, only use a single transduction frequency. This article aims to compare the two systems in principle. higher frequencies of 100-600kHz are is because the conversion of electrical energy order to elicit meaningful similarities between commonly used. to mechanical sound energy is done through the two technologies. Radar frequencies cover a wide spectrum, the use of specific piezoelectric crystals from the high frequency range of 3-30MHz Most static radar and sonar platforms are which have limited frequency modulating for the purpose of over-the-horizon radar BASIC FUNCTIONS capable of operations round the clock, thus capabilities. surveillance to the ultra-high frequency range providing a seamless and perpetual field of OF PROPAGATION of 40-300GHz – also known as the millimetre security. Radar and sonar are also capable of DETECTION SYSTEMS wavelength band – used for extremely high measuring the speed of the target in the resolution and short-range imaging. Fire following ways: Radar and sonar are sensor systems that control and missile guidance radars operate CAPABILITIES OF RADAR use the propagation of waves to detect and between these two extremes, at frequencies AND SONAR • Measuring distance and recording where localise targets. In radar, a transmitter with of 8-12GHz. the target was located a set time ago an oscillator is used to generate radio waves, Radar and sonar are capable of distance • Analysing the Doppler signal where the and a waveguide links the transmitter to the The detection range of sonar installed on measurement through the timing method, target’s motion relative to the transmitter antenna. In sonar, electrical energy is supplied large shipboard platforms varies from tens i.e. Range R = cTr/2 where c is the speed of and receiver produces a change in directly to a hydrophone array which converts of kilometres (for active sonar) to hundreds the wave, and Tr is the time taken for the frequency (see Figure 1) it to sound waves. These transmitted waves of kilometres (for passive sonar). Smaller signal to travel to the target and back. bounce off upon coming into contact with a versions of sonar are installed on helicopters Different radar transmission types are often target. The reflected waves (or the target’s and unmanned surface vehicles, and This method is considered more accurate, combined to perform a comprehensive range inherent sound in the case of passive sonar) miniaturised versions are used for unmanned although the frequency and phase of functions. For instance, static platform make their way back to the sensor system. underwater vehicles that offer situational modulation of the transmitted signal vary radar may perform a wide area search to updates in localised areas of operations. according to time. While frequency agility identify targets within a 360 degree arc For radar, the received signals normally Sonar detection can also be achieved using in radar is established and widely available, quickly. Subsequently, it executes a precise go through a low noise amplifier before the bistatic principle, although this area is still the use of several transducer frequencies tracking of targets previously identified for being down-converted to an intermediate in the developmental stage. in sonar is still in the early stage of accurate fire control. When the radar tracks frequency (Skolnik, 1982). On the other 93 development. At this stage, some sonar a target, it may vary its waveform to adapt hand, the received signals for a sonar go The spectrum of sonar frequency spans from transmissions are capable of being modulated to the target’s kinetic parameters and the through a pre-amplifier to shape the signal a few Kilohertz or less for long-range target to operate in more than one frequency environment so as to optimise its detection (front-end conditioning) prior to being sent detection to medium range of 10-50kHz due to its larger bandwidth. However, the capability. to the signal processing unit. Subsequent for target detection and classification. majority of current sonar transducers can Comparative Analysis of Radar and Sonar Principles

94

Likewise, sonar can • Amplifying the transmitted signal, if any sonar beams can either be mechanical or operate in different • Increasing the directivity of both electronic. Directivity can also be used during transmission modes to transmitted and received signals reception to cut down on unwanted noise optimise its detection • Collating, cleaning and amplifying the interference from the surroundings. and classification received signal capability for stationary Sonar utilises a deep sound channel or slow moving targets Two basic radar-transmitter configurations occasionally, where propagation may occur as well as fast moving are the self-exciting oscillator and the power with virtually no loss if the source of sound ones. Besides being amplifier. The magnetron power oscillator is is deep and the conditions are optimal. This able to achieve omni- the most common type of power oscillator is due to the sound trapping in the channel directional coverage, for radar. The klystron, travelling-wave tube with no loss at the boundaries. This sonar can operate and the cross-field amplifier are examples phenomenon is also a well-known in directed sectoral of power amplifier tubes. In the late 1990s, observation in radar, where the specific transmission to improve the solid-state amplifier started to replace atmospheric conditions can create a confined its detection capability. the vacuum tube. The choice of transmitter conduit that follows the earth’s curvature, The incorporation of an depends on the radar application. For resulting in ducting. Thus, over-the-horizon Automatic Detection instance, power amplifiers are more stable detection capabilities can be achieved. and Tracking feature in the sonar system Most modern radar systems are capable of and provide greater power at the expense of However, this phenomenon can also cause enables continuous tracking of the target for varying their S/N ratios to take into account size and portability. problems in radar performance (e.g. gaps in an accurate torpedo fire control solution. the presence of clutter, be it permanent (e.g. radar detection or detection of unwanted nearby foliage and buildings) or sporadic In the case of active sonar, transducer clutter echoes at long ranges) as well as SIGNAL AND NOISE (e.g. rain and atmospheric disturbances). The frequencies have undergone an innovation errors in range and angular measurements adaptive S/N ratio allows radar systems to where multiple frequencies may be emitted, (see Figure 3). optimise the rates of detection even under The Signal to Noise (S/N) Ratio is defined as allowing optimal performance through a adverse conditions. the power ratio between a signal (desired variety of mediums. This development may MATCHED FILTERING reading) and background noise (undesired benefit military systems in the future by The sonar equivalent of clutter is reading). All sensor systems which use the providing modes of detection suitable for When receiving a signal, a matched filter reverberation. Reverberation is caused by wave propagation have to factor in the various sea conditions. option is available in some radar applications sound waves scattering upon contact with S/N ratio in order to make meaningful where the shape of the return pulse is small objects in the sea, the water surface and calculations (see Figure 2). As sound and radio waves travel, they known. This allows for optimal reception the seabed. A region of strong reverberations suffer from a loss of intensity dependent of the desired signal by blocking out is known as a reverberation-limited A high S/N ratio is always desirable as it on the square of the range travelled. As an undesired interference. Radar pulses can also environment. Conversely, a low reverberation allows for higher rates of detection with a omni-directional transmission system loses be integrated, thus amplifying the signal condition where noise dominates is known lower probability of false alarms. All wave power quickly from a dissipation of power while cancelling out random noise. as a noise-limited environment. The sources propagation sensor systems use some form across a wide area, there is a need to add Integration can either be coherent or of noise include shipping traffic, sea state of noise limitation and signal amplification directivity to the transmitted waves to non-coherent. Non-coherent integration and the ocean’s biological system which in order to improve the S/N ratio. increase the strength of the transmitted compromises efficiency for the convenience comprises organisms such as shrimps and signal. Radar systems may use a steerable of not having to preserve phase information. fishes. Unwanted sounds and reflections can ‘Clutter’ refers to unwanted echoes in parabolic dish whereas sonar systems may use Matched filtering is also used extensively in mask a desired target from detection. electronic systems and they occur particularly a steerable beam or directed transmission. coherent active sonar. in radar. These echoes are caused by Methods of directing and steering radar and unwanted detection and interference from POWER GENERATION, a variety of environmental factors, including AMPLIFICATION AND insects, chaff and atmospheric turbulence. PROPAGATION All forms of clutter must be estimated and factored into the equation as they lower the In order to improve the S/N ratio, sensor S/N ratio significantly. Noise dominates in the systems must also include measures to absence of clutter while clutter dominates in 95 increase the power of the received signal. regions of heavy clutter interference. Radar This is done in three main ways: systems in these two scenarios are termed ‘noise limited radar’ and ‘clutter limited radar’ respectively. Figure 3. Effects of atmospheric ducting on wave propagation (Source: RF Cafe) Comparative Analysis of Radar and Sonar Principles

96

by particular elements. For instance, the 50- faster rate. Conventional wear ATTENUATION 75GHz bands are used rarely in radar as they and tear is also reduced. suffer severe attenuation due to atmospheric ‘Attenuation’ refers to the gradual loss absorption, especially under rainy conditions. Through the manipulation of in intensity of any kind of flux through a the T/R modules, AESA radar medium. Both radar and sonar suffer from The attenuation of sonar wave propagation systems are able to transmit attenuation that results from factors such bears some similarities to that of radar. multiple ‘mini-beams’ to track as range and environmental conditions. Owing to the nature of its medium, there are a large number of identified For example, as range increases, the signal problems in using sonar to propagate sound targets while maintaining a becomes weaker until it is completely hidden waves through water mediums of different wide area search beam – this by noise. temperatures, salinity and density. Such eliminates the previous need variations cause differences in the speed of for multiple radar systems Attenuation levels vary based on the sound wave propagation. (Russel, 2007). frequency of transmission and the medium through which the signal travels. It is a key At a depth of 30-100m in the open sea, Further advantages of AESA factor in determining the type of detection there is usually a marked change in ocean be cooled using silent convection, or by radar systems include low probability of method to be used. For instance, sonar is temperature, known as a thermocline. towing sonar hydrophones behind a ship detection by radar warning receivers and high used by submarines as radio waves suffer Thus, sonar can be hampered by the or submarine to reduce the effects of vessel resistance to jamming. As AESA radar systems severe attenuation underwater and become refraction of sound waves from the other noise (see Figure 5). are composed of multiple T/R modules, virtually non-functional. side of the thermocline because of the isolated failures have little effect on the difference in sound wave speeds at various The passive sonar can be compared to the entire system and this allows the AESA radar By varying frequency, radar and sonar can temperatures. An analogy would be the use of non-cooperative emissions from systems to attain high levels of reliability. make trade-offs between target resolution distortion of a drinking straw when it is commercial broadcast or communication and detection range. Higher frequency radar placed in clear water. Such variables can only signals in passive radar to detect targets of As the AESA radar systems continue to would allow for a more precise tracking of be approximated during sonar calculations. interest. In such a system, the receiver uses develop, cheaper and more power-efficient the target at shorter ranges, while lower This phenomenon is similar to radar third-party transmitters in the environment T/R modules can be expected in the market, frequency search radar would detect targets propagation through different refractive and measures the arrival time difference further reducing the cost. beyond the horizon. layers in the atmosphere. between a signal obtained directly from the transmitter and one which is obtained via Sonar systems also have equivalent array An example of attenuation arising from PASSIVE SYSTEMS reflection from the object. used for receiving signals. The sonar array environmental conditions for radar is rain consists of multiple hydrophones which fade, as illustrated in Figure 4. Higher add signals from a desired direction while Passive sonar systems are used extensively ACTIVE ARRAY RADAR frequencies of radio waves (e.g. those subtracting signals from other directions by mines and submarines. Passive sonar is above the range of 11GHz) suffer most from SYSTEMS (Chapple, 2008). Hydrophones are most able to determine the range and bearing of attenuation due to rain fade. Specific bands commonly arranged in the format of a line an acoustic target without giving away the of frequency also suffer from attenuation Modern Active Electronic Scanned Array array and cannot be steered electronically. location of its source. Once an acoustic target (AESA) radar systems as shown in Figure 6 has been identified using provide great flexibility as they are capable broadband detection, the of imitating a variety of radar antennas sonar vessel may use narrow according to the situation (Hommel and band analysis to identify Feldle, 2004). They are composed of several the different frequencies Transmit/Receive (T/R) modules which are making up the target’s programmed to operate in tandem. emitted sound. Hence, the type of engine and craft can AESA radar systems are able to generate be determined. and aim the radar transmission beam and reception path electronically through the It is most useful to reduce constructive and destructive interference 97 the vessel’s own noise. between the T/R modules. This eliminates the This can be achieved by need for physical movement of the radar in Figure 6. An AESA radar employing nuclear reactors order to carry out scanning, allowing the AESA in submarines which can (Source: Bahrat Rakshak) Figure 4. Rain fade (Source: Dream Satellite Television) radar systems to scan surroundings at a much Comparative Analysis of Radar and Sonar Principles

98

SYNTHETIC APERTURE CONCLUSION BIOGRAPHY RADAR AND SYNTHETIC Despite inherent differences between radar APERTURE SONAR and sonar, there are similarities in the nature Teo Seow Khye is Head (Sensor Systems) and is responsible for the of their sensor systems which lead to overlaps operations and engineering support for non-platform sensor systems Synthetic Aperture Radar (SAR) is a type of in their applications and technologies. Cross of the Singapore Armed Forces (SAF). He was part of the team for the radar which collates multiple radar images radar and sonar technical competencies can Republic of Singapore Navy (RSN) Missile Corvette Project and the the RSN to yield a single image of high resolution. be seen as synergistic, and innovations in one Frigate Project which attained the Defence Technology Prize Team Award The quality of the resultant image is better field are likely to affect and benefit the other. in 1990, and the Team (Engineering) Award in 2007 respectively. Seow than what is achievable through standard Hence in DSTA, the radar and sonar specialists Khye received the Public Service Commission Local Scholarship Award and conventional means. Either a single antenna are put together in the same competency graduated from the National University of Singapore (NUS) with a Master is mounted on a moving platform (e.g. an community to update one another with their of Science (Electrical Engineering) degree in 1990. He further obtained airplane or spacecraft) to illuminate a target respective system developments. a Master of Science (Management of Technology) degree from the scene or many low-directivity small stationary Massachusetts Institute of Technology, US under the DSTA Postgraduate antennas are scattered around the target Scholarship in 1993. area. The many echo waveforms received REFERENCES at the different antenna positions are post- processed to resolve the target. Synthetic Chapple, P. 2008. Automated Detection Aperture Sonar (SAS) is analogous to SAR and and Classification in High-resolution Sonar shares the same basic principle – the forward Imagery for Autonomous Underwater Vehicle Neo Chin Sim is a Principal Engineer (Systems Management) and heads the motion of the platform is used to synthesise Operations. Defence Science and Technology Torpedoes Systems Management programme. He has overseen projects a long antenna which in turn improves the Organisation Publications Online. in the areas of underwater sonar, torpedoes and underwater signature azimuth resolution (Hagen and Hansen, management. Chin Sim obtained his Bachelor of Engineering (Electrical and 2008). Hagen, P.E. and Hansen, R.E. 2008. Electronic Engineering) degree and a Master of Science (Management of Synthetic Aperture Sonar Challenges. Hydro Technology) degree from NUS in 1991 and 2001 respectively. The differences between radio waves and International 12(4):26-31. sound waves give rise to a different set of challenges in SAS as compared to SAR. For Hommel, H. and Feldle, H. 2004. Current example, in order for the sonar to function Status of Airborne Active Phased Array (AESA) as a synthetic aperture, the position of the Radar Systems and Future Trends. Paper sonar at each ping must be determined presented at the European Radar Conference, Solomon Sia Hou You was awarded the DSTA Undergraduate Scholarship with great accuracy. Hence, the underwater Amsterdam, 11-15 October. in 2008 and is currently studying Electrical and Computer Engineering at vehicle carrying the sonar has to be equipped Carnegie Mellon University, US. He was on an internship at DSTA from March with a highly accurate navigation system. Minkoff, J. 1991. Signals, Noise and Active to May 2010 where he learnt about the different radar systems fielded by In addition, factors such as electronic and Sensors: Radar, Sonar, Laser Radar. John the SAF. ambient noise become more important as Wiley and Sons Ltd. range increases, and multipath reverberation may also increase to the point that the Russel, M.E. 2007. Future of RF Technology maximum effective range of the sonar may and Radars. Paper presented at 2007 IEEE be limited by reverberation in shallow water. Radar Conference, Boston, MA, USA, 17-20 While these issues are also faced by SAR, April. solutions such as the use of interferometry techniques have been developed to resolve Skolnik, M.I. 1982. Introduction to Radar multipath problems. As the engineering Systems. McGraw-Hill. challenges associated with SAS processing have proven to be more difficult to overcome than those associated with SAR, the latter is now more widely used. 99 Planning and Designing ABSTRACT There are many schools of thought on data centre design and implementation. Each has its own benefits and challenges. Data Centres Organisations will have to consider the specific business needs to come up with their requirements and designs for their data centres. Usually, these data centres are based on unique design concepts, taking into account industry best practices and infrastructure constraints. This article shares some of the design considerations and implementation approaches for data centres.

Yee Keen Seng Wu Xianghua Lim Hwee Kwang Wong Marn Chee Ang Choon Keat Planning and Designing Data Centres 102

on the number of data centres and their It can be seen that the option of building Besides analysing the historical IT rate of INTRODUCTION locations. four data centres requires an addition of growth, it is useful to review the initiatives three sets of M&E systems infrastructure and that the organisation has carried out over the Data centres house two of an organisation’s For operational efficiency and cost IT common services infrastructure. Space will last few years and study its envisioned state most valuable assets – data and IT systems. effectiveness, it seems attractive to also be taken up by the IT common services of IT maturity. This will help to determine Well-designed data centres based on a consolidate IT systems into a single data infrastructure built at each data centre, the future initiatives of the organisation, robust strategy will enhance the availability centre. When space, power and cooling effectively reducing the total usable space for normalise the spikes and troughs of the IT and reliability of the IT services to the end capacities are optimised, developmental the business IT systems of the organisation. requirements and give a better projection of users. This allows organisations to respond and operational costs of the data centre With four data centres, the M&E systems and the space required for the data centres. efficiently and effectively to business are reduced. Centralising the IT systems also IT common services infrastructure at each demands and challenges, which is a critical minimises the human resources used for data centre may be smaller in capacity, but For example, if the organisation is success factor for organisations. managing the IT systems. When operations the total cost of ownership will be higher as planning to implement projects such as are conducted in close proximity to one compared to having a single data centre. knowledge management or video content This article discusses some of the basic but another, processes can also be streamlined management, the data centre space will key considerations for planning and designing more effectively. However, if the single data To balance resiliency and cost, most increase significantly due to large storage data centres, which are dependent on the centre fails, the entire organisation may be organisations will have two to four requirements of these projects. On the other organisation’s business needs, tolerance for brought to a halt. data centres to support their business hand, if the organisation is embarking on an failures and availability of budget. However, requirements, and a reliable disaster recovery IT consolidation or virtualisation path, there certain aspects of the data centre such as Distributing the IT systems over multiple plan for these data centres. may be a reduction in space requirements for total space and structural loading cannot be data centres at different sites provides the IT infrastructure over time. changed without major redevelopment of required resiliency to the organisation. the data centre and disruptions to day-to- The organisation will not lose all of its SPACE CONSIDERATIONS day operations once it has been built. Thus, operational capabilities should one of its ARCHITECTURAL The organisation has to determine the it is critical that these aspects are given due data centres fail. Certain business units or CONSIDERATIONS amount of space required for each data consideration during the planning phase. functions may not be available but the rest of the organisation can continue with their centre. Undersizing the data centre may result The practice of architecture is primarily in it running out of space quickly and new NEED FOR DATA CENTRE daily operations. However, multiple data concerned with the principles of durability, centres require IT infrastructure to be set data centres may need to be built frequently. functionality and human comfort. While STRATEGY up at multiple sites. More manpower may This may go against the organisation’s data the first two principles can be quantitatively be required to conduct and coordinate centre strategy and incur additional costs. satisfied, the solution to the third is not as Some organisations use commercial data operations at the various sites which would On the other hand, oversizing may lead to straightforward. Designing a building for centres to host their IT systems while others increase the developmental and operational wastage of resources and space that could be servers presents a unique challenge as it may prefer to build their own data centres for the costs. put to better use. not be intuitive to the architect, who designs purpose of security and control, among other spaces for humans instead of machines. reasons. As part of the organisation’s overall Table 1 illustrates the various considerations It is relatively easy for an organisation to IT masterplanning, a data centre strategy in deciding between building a single data know what it needs now and probably for In a data centre, optimal working must be developed, including the decision centre and four smaller data centres. the next three years. However, it is very environment of computer servers takes difficult to forecast the requirements for the precedence over human comfort. Thus, next 10-15 years given the rapid changes in M&E requirements drive the design of the Components Single Data Centre Four Data Centres technology and business. data centre and its organisational principles. Assumed Assumed However, the needs of operators should Building Construction similar cost similar cost One possible way of estimating the space not be neglected. The architectural design requirement is to look at the historical IT serves to complement the layout of the data Mechanical and Electrical (M&E) Systems One Set Four Sets growth of the organisation and assume Infrastructure centre in ensuring that circulation patterns the same rate of growth for the next 10-15 are logical and conform to building codes and IT Common Services Infrastructure years. However, this approach has a potential regulations. For example, circulation spaces One Set Four Sets (Transmission/Network, Security, Storage, etc) pitfall – the focus and initiatives undertaken can be designed to allow natural lighting – by the organisation in previous years may this can relieve operators of the stress from Usable Space for Business IT Systems differ significantly from those of the next few working in a hermetic environment. 103 (Assuming IT Infrastructure Common Services X – Y X – 4Y years. In that case, there may be undersizing take up Y rack space) or oversizing of the data centre space It may be useful to separate administrative Table 1. A comparison of infrastructure requirements requirements. areas from data centre rooms as they have Planning and Designing Data Centres

104

vastly different spatial requirements. While modification is made to the building Tier Requirement Tier 1 Tier 2 Tier 3 Tier 4 the data centre rooms require a hermetic structure. and closed environment with large slab- 2 2 to-slab heights, the administrative areas Today, a rack fully occupied with equipment Power Supply to Data Centre 1 1 (One Primary, (Both should be open to natural lighting without such as blade servers and storage area One Backup) Primary) large vertical spaces. One possible way networks may weigh 600-1,000kg. The of resolving the different requirements structure must be adequate for supporting M&E Systems Supporting IT Equipment (e.g. Generators, N N+1 N+1 2(N+1) is to consolidate all the administrative a loading of approximately 10-13kN/sqm. Chillers, UPS, CRAC Units) areas on a separate floor or in a wing. This This is in contrast to the designed loading of Downtime Required during strategy will also aid in compartmentalising 5kN/sqm adopted in many buildings. In Yes Yes No No the building according to use, security the future, IT equipment is expected to be Maintenance requirements and susceptibility to fire heavier due to the incorporation of more Data Centre Availability 99.67% 99.74% 99.98% 99.99% hazards. electronic components. This means that future data centres should have a structural Data Centre Downtime Per Year ~28.8hrs ~22.0hrs ~1.6hrs ~0.4hrs STRUCTURAL LOADING loading higher than 13kN/sqm. New data centres built today are generally designed to Medium High Very High CONSIDERATIONS Relative Cost* Low support 15-25kN/sqm in structural loading. (~22 Tier 1s) (~2 Tier 2s) (~2 Tier 3s)

The required loading on the structure of the * Relative Cost is a general industry assessment and not part of the TIA-942 Standards. data centre is an important consideration for POWER AND COOLING Table 2. Data centre availability tiers by the Uptime Institute the design of the data centre building. This CONSIDERATIONS is normally specified in terms of distributed requiring power and cooling capacities of downtime for a data centre. Table 2 floor loading and expressed in kilo Newtons M&E systems can easily make up more than approximately 10kW per rack. Thus, there provides a summary of the four data centre per square metre (kN/sqm). The building 60% of the total developmental cost of is a two or three-fold increase in power availability tiers. structure must be adequate to support the a new data centre and is thus a major cost and cooling requirements for racks of blade weight of the equipment required in the component. M&E systems include electrical servers as compared to racks of rack-mount Table 2 shows the various configurations data centre. substations, chillers, backup generators, servers. for the design of M&E systems e.g. N+1. uninterrupted power supplies (UPS) and ‘N’ refers to the minimum number of M&E Structural loading considerations are critical computer room air-conditioning (CRAC) To determine the right size of the power equipment required to support all the for the safety of the entire building where units. Careful consideration of the power and cooling requirements for the data operations of a data centre and ‘+ X’ means the data centre resides. Overloading the and cooling requirements is required to centre, it is important for the organisation that there are X spare systems to back up building structure may eventually render the implement M&E systems of an appropriate to set a direction for the equipping of IT the main systems. It is very expensive to set building unsafe for occupancy. In the worst size to avoid wastage of resources. infrastructure. The organisation can chart up a 2(N+1) configuration at Tier 4. M&E case scenario, it may even cause the building the power and cooling requirements in systems contribute significantly to the cost to collapse. Hence, it is important to consult The power and cooling capacities to cater phases over the life cycle of the data centre. of a data centre. Unless there are very critical a qualified structural engineer before placing for a data centre depend on the type of IT However, when the data centre is first built, operations to support, most organisations an exceptionally heavy piece of equipment equipment to be deployed. A single rack space needs to be provisioned for the amount will settle for a lower data centre requirement in the data centre during the operational can pack more blade servers providing of M&E systems to support the eventual or just the N+1 configuration. phase. more computing resources per rack than power and cooling requirements. rack-mount servers. However, blade servers The N+1 configuration means that There are several methods to manage the generally require higher power and cooling REDUNDANCY maintenance of the M&E systems will scenario where the weight of equipment capacities than the rack-mount servers. More be carried out one at a time. The single is heavier than the structural loading. One space may be required for cooling which CONSIDERATIONS spare system will fill the gap while the way is to load each rack with less equipment in turn reduces the advantage of space maintenance is ongoing. This is feasible if Although not mandatory, most of the and increase the number of racks used. An efficiency when adopting blade servers. the data centre is small. However, if the data centres today adopt the TIA-942 alternative is to use spreaders – usually a data centre is large with numerous units of Telecommunications Infrastructure Standards metal plate strong enough to hold the rack For instance, a standard IT rack can pack M&E equipment, the entire maintenance for Data Centre in their designs. Four data and equipment – to distribute the weight about 32 full-height blade servers or 64 half- process may spread over days, weeks or even centre availability tiers, defined by the 105 over a wider area. These two methods will height blade servers, which require power months since only one system can undergo Uptime Institute, are incorporated into reduce the usable space for IT in the data and cooling capacities of approximately maintenance at a time. centre. A qualified structural engineer will 20-25 Kilowatts (kW) per rack. A similar the TIA-942 guidelines which specify the also need to be consulted before any rack can pack about 20 rack-mount servers, redundancy and tolerable unplanned Planning and Designing Data Centres

106

Most importantly, the N+1 configuration data centres while equipping M&E systems. workflows. Essentially, form should fit means that the data centre cannot afford Oversizing results in higher developmental SECURITY AND function. For example, areas outside servers to fail during the maintenance of the M&E costs when purchasing more and bigger M&E OPERATIONAL should be created as service corridors to systems. During maintenance, when one of equipment, as well as higher operational CONSIDERATIONS house CRAC units. Thus, CRAC service the M&E systems shuts down, the spare M&E and maintenance costs. During operations, personnel do not need to have physical system will be activated to take on the load higher energy consumption will be required A well-designed data centre will consider access to the servers during maintenance of of the inactive system. If the spare M&E to power the oversized M&E systems not only the availability and reliability of the CRAC units. Printers and backup devices system fails at this point in time, there will not performing at an inefficient level. The the data centre facilities, but also security should be housed in separate rooms from be another M&E system available to activate higher energy consumption to support an and operational requirements. Crucial the server areas. Such measures enhance and support the load from the IT systems. oversized M&E system is a waste of energy. organisational information is housed in the operational workflows as the personnel Hence, some IT systems will be affected. The organisation should adopt a modular data centre. Thus, it is imperative that the servicing different subsystems do not need approach and equip the M&E systems when data centre is kept secure to prevent data to share physical space and obstruct one To overcome these potential shortcomings, required. leakage and sabotage. another. many organisations design their data centres with N+2 or N+3 configurations. In case With advancement of technologies and Multi-layer security implementation is Security measures and operational efficiency a failure occurs during maintenance, the the manufacturers’ increasing focus on recommended. All personnel at the data may not always be aligned. Depending on second or third spare system can be activated being environmentally friendly, most M&E centre including staff and visitors must how crucial the data centre is to the to support the power and cooling loads. systems and IT systems are designed for be screened and cleared by the security organisation, different levels of security energy efficiency. For example, linking implementation may be needed. In all cases, department. Access to the data centre temperature sensors to smart CRAC units to the data centre design should give due GREEN DATA CENTRE should be restricted such that only authorised vary the cooling capacity of the data centre consideration to the operations of the data CONSIDERATIONS personnel are allowed to enter the areas can lead to substantial savings in energy centre. Where there are conflicts, security that they are required to work in or visit. All consumption. Rainwater can be harvested usually takes precedence over operational There is an increasing focus for data centres equipment in the data centre should also to cool plant systems so as to conserve efficiency. to be green. A green approach to design have security locks or equivalent safety water. These technologies may incur slightly ensures that a building is environmentally measures to control physical access to them. higher data centre developmental costs friendly and resource-efficient throughout but great savings will be reaped in the life Activities in critical operational areas of the VIRTUALISATION its life cycle from construction to demolition. cycle operational costs. Thus, they should data centre should also be monitored and Going green brings tangible savings to the In addition to data centre design be deployed for the data centre as far as recorded. organisation not only through reduced considerations, IT strategies and possible. power consumption, but also in the form In addition to these basic security implementation plans also play an important of greater durability and economical usage role in enhancing the efficiency of a data The standards for green data centres are measures, data centres should also consider of natural resources. In addition, a green centre in terms of space, power and cooling evolving, spearheaded by organisations like implementing more advanced technologies, environment also benefits the occupants utilisation. the Green Grid at the international level as such as active radio frequency identification of the building. For example, maximising well as the Building and Construction Authority and sensors, to track the actual location of natural lighting in administrative office areas Most servers are operating at a utilisation (BCA), Infocomm Development Authority of people in the data centre. It should come or using low volatile organic compound rate of only approximately 15-20% when Singapore (IDA) and SPRING Singapore on the with an alarm system to detect people who paints can lead to a more conducive working running a single business application. Most local front. SPRING Singapore has published the have entered areas without authorisation. environment and increased productivity. servers also use approximately 40-50% of “Singapore Standard for Green Data Centre – Some consultants may recommend triple their power capacities when idle. An Energy and Environment Management Systems authentication access measures using cards, Energy efficiency in server rooms can often organisation can address this inefficient use (SS 564 : 2010)” in January 2011, enabling an passwords and biometrics. However, the be achieved with simple and inexpensive of IT resources using virtualisation. With organisation to plan, build, operate and maintain actual additional security achieved should measures. For example, installing blanking virtualisation, multiple business applications a green data centre, by establishing the systems be measured in the context of multi-layer panels in racks and arranging the racks into can run on a single physical server. This and processes to manage and improve the data security architecture. The cost incurred in hot and cold aisles will improve the airflow increases utilisation and power efficiency centre’s energy performance. These standards implementation, security administration dynamics and raise the cooling efficiency. In of servers. At the same time, it reduces the should be considered for adoption if they do and operations should be balanced with the administrative areas, natural lighting can be number of servers deployed, leading to fewer not compromise security or cardinal business additional security achieved. used to lessen lighting loads. idle servers expending data centre power and requirements. 107 Operational workflows have to be cooling capacities, as well as better utilisation It is important to determine the right size formulated around security policies and of space in the data centre. of M&E systems to ensure requirements are implementation. The internal data centre met adequately. In anticipation of future has to be designed to meet these operational needs, some organisations tend to oversize Planning and Designing Data Centres

108

emergence of future technologies with lower PUTTING IT TOGETHER power and cooling requirements. However, it Wu Xianghua is an Architect (Building and Infrastructure). He is working is more likely that smaller footprints would on the future expansion of the SAFTI Military Institute. Xianghua received Organisations can consider starting with two lead to more equipment packed into the his Bachelor of Architecture degree with First Class Honours, with a minor data centres. It is not advisable to have just data centre, leading to increased demand for in Architectural History from Carnegie Mellon University. He also obtained one data centre as its failure will affect the power and cooling capacities. Fortunately, numerous awards for his design work and for being the top graduate in the survivability of the organisation. Cardinal not every piece of equipment has high power architecture programme. IT systems that house critical business and cooling requirements e.g. network applications and data can be split between equipment. Based on industry trends, many two data centres. This will maintain at least organisations are designing their data 50% of the operations if one of the data centres for an average power utilisation rate centres fails. A modular build-up approach of 10kW per rack for current requirements can be used to size data centres. and possibly provisioning for 20kW per rack Lim Hwee Kwang is an Assistant Director (Infocomm Infrastructure). for future requirements. Working with MINDEF and the SAF, he strategises and develops the M&E systems should be built and equipped masterplan for key Infocomm infrastructure ranging from data centres to based on anticipated requirements for the CONCLUSION messaging infrastructure. He leads, directs and manages the development, next five years. Sufficient space has to be implementation and engineering support of information infrastructure. catered for future expansion of the data Given the large investments in data centres, Hwee Kwang co-developed the Civil Service-wide Secure Messaging System centre. Assuming that future technologies organisations should plan their data centres with the Infocomm Development Authority of Singapore. He also designed have a smaller requirement for space as properly, consider their available budget and and implemented enterprise-wide Public Key infrastructure and compared to current technologies, it is tolerance for failures and disruptions. The authentication solutions. Hwee Kwang holds a Master of Science likely that future equipment can fit into the organisation must anticipate its immediate (Information Security) degree from Royal Holloway, UK and a Master of space reserved for expansion. This reserved and long-term business needs and align them Science (Management of Technology) degree from NUS. He further attained space can be used as temporary offices or with its IT strategy. the Chief Information Officer Certificate as a Top Distinguished Graduate storage rooms until the need for data centre in the National Defense University, US in 2007. expansion arises. REFERENCES The challenge remains in the accurate estimation of power and cooling capacities. Telecommunications Industry Association. Wong Marn Chee is a Senior Engineer (Infocomm Infrastructure) and the The increasing emphasis on being 2010. TIA-942 Telecommunications Head of Defence Technology Tower Computer Centre. He oversees the environmentally friendly may lead to the Infrastructure Standards for Data Centres. development, engineering support and IT operations of the computer centre. Marn Chee plays a key role in the development of Enterprise Server Farms for MINDEF and the SAF. He has many years of experience in designing, BIOGRAPHY architecting, implementing and managing the engineering support for the server farms and their common network services. Marn Chee received his Bachelor (Electrical Engineering) degree with Honours from Nanyang Yee Keen Seng is a Senior Engineer (Infocomm Infrastructure). Technological University in 1995. He is currently involved in planning, designing and developing the first purpose-built data centre for the IT systems of the Ministry of Defence (MINDEF) and the Singapore Armed Forces Ang Choon Keat is a Senior Engineer (Building and Infrastructure) and (SAF). He oversaw several best-sourcing projects including the works on planning, designing and managing protective building and provision of shared services and end-user IT support to MINDEF and the SAF. infrastructure developments. He is a registered Professional Engineer (Civil) Keen Seng served in the SAF Chief Information Officer Office where in Singapore. Choon Keat was previously Manager (Plans) at the Defence he managed the development and governance of the SAF Enterprise Research and Technology Office where he was involved in the management Architecture Framework as well as pioneered the Ops-Admin Systems of the MINDEF Research and Technology portfolio. He graduated with a Integration initiative. Keen Seng is certified by Enterprise Products Master of Engineering (Civil Engineering) degree from Imperial College Integration (EPI) as a Data Centre Professional and a Data Centre Specialist. London, UK in 1999 under the DSTA Undergraduate Scholarship. He further He holds a Bachelor of Science (Information Systems and Computer Science) obtained a Master of Science (Operations Research) degree from Columbia 109 degree from the National University of Singapore (NUS). University, US in 2005 under the DSTA Postgraduate Scholarship. Choon Keat is currently pursuing a Doctor of Philosophy (Management of Technology) degree from NUS. Effects of ABSTRACT While there are documents providing guidelines on ground Degree of Saturation shock loading on underground structures, the application of such guidelines in a tropical country like Singapore needs to be re-examined. Singapore experiences high temperature and on Ground Shock abundant rainfall throughout the year. Therefore, the soil moisture surrounding an underground structure above the ground water table alternates between dry and wet conditions. This article summarises the findings of a series of small-scale field tests to investigate the effects of the degree of saturation on ground shock. These findings presented will aid the development of technologies to mitigate ground shock effects on underground structures in the future.

Leong Eng Choon Sarma Anand Lim Chee Hiong Effects of Degree of Saturation on Ground Shock 112

INTRODUCTION GROUND SHOCK AND UNDERGROUND Going underground is an effective way of avoiding enemy detection and reducing the STRUCTURES risk of physical attacks. Finding deeply buried facilities remains as one of the greatest Ground shock is the loading of the ground technical challenges in warfare (Streland, due to the explosion of bombs on or 2003). The history of using underground within the ground and is the main threat structures dates back to the Second Century to underground structures. There are three AD when early Roman Christians hid key parameters that have to be considered in catacombs – which are underground for the design of an underground protective structures dug from porous rock – outside of structure: ground shock, structure and the Rome to escape persecution. From 1929 to ground in which the structure is installed. 1940, the French constructed a 235-kilometre- However, these three parameters are not long subterranean system called the Maginot mutually exclusive. Ground shock depends on Line to protect France from German the source (i.e. bomb) and the transmitting Figure 1. A paraboloid fibreglass self-contained shelter model incursions (Allcorn, 2003). During World War medium (i.e. ground). Ground shock intensity (Source: Adapted from McCarthy, 2002) Two, the Germans used underground mines on an underground structure is the most to house military industries and missile silos. difficult to estimate (Baylot, 1992). covered by sand was able to withstand TM 5-855-1 (1986) as well as by Drake, Frank Most research efforts have concentrated In the Cold War era, Sweden built a much higher overpressures as compared to a and Rochefort (1987) and Baylot (1992). on improving the strength of underground number of underground defence facilities. cylinder covered by clay. Baylot (1992) structures. For example, Gautam and Pathak Since then, the US has also built several suggested that there are two options The peak pressures for contained explosions (1997) described an underground blast- underground facilities. The construction of available to designers of underground in various soil types as a function of scaled resistant structure made of lightweight an underground command post at Raven structures: (a) to replace the in-situ soil distance is shown in Figure 2. Besides soil steel sheets that can withstand a dynamic Rock Mountain, Pennsylvania, for the with sand or (b) to use the in-situ soil. type, the degree of saturation plays an loading of 12psi. McCarthy (2002) described US military was completed shortly after Although the former option provides a important role in determining peak pressures paraboloid fibreglass self-contained President Eisenhower’s inauguration in 1953. more favourable condition, it is more (Leong et al., 2003; 2007). As shown in Figure structures (as illustrated in Figure 1) that In 1962, an underground shelter built to expensive and it may not be feasible in 2, corresponding higher peak pressure can withstand overpressures of 15-20psi. house the legislative branch was completed some cases. Nevertheless, free-field stresses occurs as the soils get increasingly wet (as However, very little attention has been in West Virginia. During the 1950s and can be used to estimate the loading on the shown by curves 2 and 3). It is recommended paid to the ground conditions in which the 1960s, a blast-proof vault was constructed structure. This observation was discussed in that underground shelters should be in a mountain in the US to store money that structure is installed. This article highlights would keep its economy running in the event the importance of ground conditions in of a devastating attack. the installation and design of underground structures. To date, the number of Continuity of Government underground facilities in the The Technical Manual on Fundamentals of US is between 50 and 100. Linger, Baker Protective Design for Conventional Weapons and Little (2002) contended that going (TM 5-855-1) recognises that ground shock underground reduces the risk of critical intensity can vary widely for different soil infrastructure disruptions from terrorist types. Experiments conducted by Dorris attacks. The ground and its moisture (1965) and Walker, Albritton and Kennedy conditions are important factors to consider (1966) examined the response of a steel in the construction of underground structures cylinder (with the dimension of four inches as they determine the structural loads in diameter, 18 inches in length, and 1/16 inch in height) placed at several depths of significantly. This article focuses on the 113 ground moisture conditions around an cover ranging from dense dry sand and stiff underground structure in response to a clay to static and dynamic overpressures. conventional weapon load. The experiments showed that the cylinder Figure 2. Peak stresses from contained explosions in various soils (Source: Adapted from TM 5-855-1, 1986) Effects of Degree of Saturation on Ground Shock

114

Figure 4. Relationship between peak pressures and distance using dimensionless parameters

Figure 3. Schematic view of small-scale explosion tests in partially saturated soil (Source: Adapted from Leong et al., 2007)

installed outside of flood-prone areas or wave which travels through the soil can be above the ground water table (The Survival visualised as travelling through a three-phase Center, 2011). However, in a tropical country medium: soil particles (i.e. solid), water and like Singapore with abundant rainfall air. Dry soil consists of soil particles and air Figure 5. Schematic diagram of a buried cylindrical silo made of throughout the year, the ground moisture while completely saturated soil consists of soil reinforced concrete and fitted with a barrier of PVC pipes (Source: levels are usually high. particles and water. This visualisation can be Adapted from Kobielak, Krauthammer and Walczak, 2007) extended to scenarios where the explosion A joint research programme between DSTA wave travels through only air or water. wet soils. Therefore, the field tests revealed basis to develop technologies to mitigate and Nanyang Technological University two observations. First, curves for different ground shock. For example, it can explain the (NTU) studied the effects of the degree of The peak pressures for explosions in air degrees of saturation will be similar to those observations in Kobielak, Krauthammer and saturation on free-field peak pressures using (TM 5-855-1, 1986) and in water (Cole, 1965) for different soil types given in TM 5-855-1. Walczak (2007), Würsig, Greene and Jefferson small-scale field tests (Leong et al., 2003; are known. From Figure 4, it can be observed Second, as the soil becomes increasingly (2000) as well as Leighton, Richards and 2007). One series of tests used one, four and that there is a merger between the curves saturated, the dimensionless peak pressure White (2004). Kobielak, Krauthammer and 10 kilograms of plastic explosive known as depicting an explosion in air and in water. will approach the curve depicting an Walczak (2007) conducted tests to investigate Pentaerythritol tetranitrate (PETN) at various Furthermore, all the curves converge at low explosion in air or water i.e. from curve 5 to the effect of a barrier consisting of a row of depths (see schematic diagram in Figure 3). values of scaled distances and spread out at curve 1. The reduction in dimensionless peak empty polyvinyl chloride (PVC) pipes on the The test results show that peak pressures high values of scaled distances. Peak pressures pressure becomes more pronounced at larger response of a buried silo to an underground increase at higher degrees of saturation. The from field tests conducted in unsaturated dimensionless scaled distances. explosion (see Figure 5). They observed that 115 results are illustrated more clearly in Figure and saturated soils are plotted on separate the barrier was effective in reducing the 4 which shows the re-plotting of Figure 2 lines. In addition, the readings of peak Knowledge of the effects of the degree peak pressures on the buried silo, and the using dimensionless parameters of peak pressures from field tests conducted in of saturation on the dimensionless peak reduction increased as the distance from the pressures and scaled distances. The explosion unsaturated soils were lower than that of pressure is useful in providing the theoretical explosion to the barrier increased. The barrier Effects of Degree of Saturation on Ground Shock

116

can be visualised as a highly unsaturated zone of the soil with large voids before the CONCLUSION buried silo and peak pressures. In Figure 4, Hammer curve 1 represents the condition without the Underground structures are important barrier and curve 4 represents the condition protective elements in national defence. Piling Barge with a barrier. Therefore, it can be seen that The design of underground structures to there is a reduction in peak pressures as withstand ground shock is improving with scaled distances increase. research and development. In this article, the research on the effect of soil moisture Würsig, Greene and Jefferson (2000) studied conditions on peak pressures provides Pile the mitigation effects of an air bubble curtain another perspective to those outlined in 25m Bubble Curtain in reducing underwater noise caused by TM 5-855-1 (1986). This new perspective percussive piling (see Figure 6). The air bubble enables a number of observations to be curtain was found to be effective in reducing explained and quantified. DSTA and NTU sound transmission through water due to are working together to understand the density mismatch as well as the combined fundamental behaviour of wave propagation Air Pipe on Sea Bed reflection and absorption of sound waves through different soil types and degrees of by air bubbles. In Figure 4, the curves are saturation. This understanding will lead to bounded by the upper peak pressure curve a better estimation of the peak pressures Figure 6. Schematic diagram of a bubble curtain used to reduce pile-driving noise (Source: Würsig, Greene and Jefferson, 2000) due to an underwater explosion. Therefore, which can be applied to develop technologies it can be expected that an underwater for mitigating ground shock effects on explosion in water containing air bubbles will underground structures. result in a lower peak pressure curve.

Leighton, Richards and White (2004) offered an explanation of the purpose of bubble nets created by humpback whales. It was observed that humpback whales create circles of air bubbles around their prey by emitting air through their blowholes. This action is accompanied by feeding calls lasting one to two minutes. It is postulated that the sound will help to gather and trap the fish for feeding (see Figure 7). Therefore, it can be inferred that keeping the surrounding soil around an underground structure in an unsaturated condition creates a barrier and prevents ground shock from reaching the underground structures.

Figure 7. Schematic diagram of a humpback whale creating a bubble net for feeding 117 Effects of Degree of Saturation on Ground Shock

118

Leong. E.C., Anand, S., Cheong, H.K. and Lim, REFERENCES C.H. 2007. Re-examination of Peak Stress BIOGRAPHY and Scaled Distance Due to Ground Shock. Allcorn, W. 2003. The Maginot Line 1928-45. International Journal of Impact Engineering, Illustrated by Vanelle, J. and Boulanger, V. 34(9):1487-1499. Leong Eng Choon is an Associate Professor at Nanyang Technological Osprey Publishing. University (NTU). He chairs the Technical Committee for Civil Works and Linger, D.A., Baker, G.H. and Little, R.G. 2002. Geotechnics and is the Deputy Chair for the Singapore Accreditation Council Baylot, J.T. 1992. Parameters Affecting Loads Applications of Underground Structures (SAC) Committee for Inspection Bodies. Eng Choon is also a member of the on Buried Structures Subjected to Localized for the Physical Protection of Critical Building and Construction Standards Committee and is a Technical Assessor Blast Effects. Technical Report, Waterways Infrastructure. Paper presented at the North for the SAC Singapore Laboratories Scheme. His research interests include Experiment Station. American Tunneling Conference 2002, soil dynamics, unsaturated soil mechanics, numerical modelling, foundation Seattle, Washington, 18-22 May. engineering, laboratory and field tests. He is involved in research projects Cole, R.H. 1965. Underwater Explosions. New ranging from the dynamic properties of Singapore soils to slope repair York: Dover Publications. McCarthy, W.W. 2002. Self-contained and stabilisation technology in Singapore. Eng Choon obtained his Doctor Underground Shelters for Pre-positioned of Philosophy (Geotechnical Engineering) degree from the University of Dorris, A.F. 1965. Response of Horizontally Deployment for Defense against NBC Western Australia in 1991. Oriented Buried Cylinders to Static and Warfare. Paper presented at the NBC Defense Dynamic Loading. Technical Report, Collective Protection Conference Colpro 2002, Waterways Experiment Station. Orlando, Florida, 29-31 October. http://nsrdec. natick.army.mil/jocotas/ColPro_Papers/ Sarma Anand is a Principal Engineer (Building and Infrastructure) and Drake, J.L., Frank, R.A. and Rochefort, M.A. McCarthy.pdf (accessed 1 March 2011) Programme Manager. He is involved in masterplanning, engineering design 1987. A Simplified Method for the Prediction and project management of airforce infrastructural works, which include of the Ground Shock Loads on Buried Streland, A.H. 2003. Going Deep: A System developing aircraft runways and specialised facilities. Anand also researches Structures. Proceedings of the International Concept for Detecting Deeply Buried Facilities and manages technological development in geotechnical engineering, Symposium on the Interaction of from Space. Research Report, Air War College, working on the aspects of liquefaction and soil improvement. He was a Conventional Weapons with Structures, Air University, USA. member of the Protective Technology Research Team which clinched the Mannheim, West Germany. Defence Technology Prize Team (R&D) Award in 2006. Anand is a certified The Survival Center. 2011. http://survival professional civil engineer, and a Fellow at the Singapore Institute of Gautam, C.K. and Pathak, R.C. 1997. Design center.com/UG.html#Under%2 Arbitrators. He also obtained a Doctor of Philosophy (Ground Shock) degree of Blast Resistant Structure. Defence Science Ground (accessed 1 March 2011) from NTU under a Defence Technology Training Award. Journal, 47(2):139-148. TM 5-855-1. 1986. Fundamentals of Protective Kobielak, S., Krauthammer, T. and Walczak, A. Design for Conventional Weapons. U.S. 2007. Ground Shock Attenuation on a Buried Department of the Army. Lim Chee Hiong is Director (Building and Infrastructure). He played a leading Cylindrical Structure by a Barrier. Shock and role in the development of Singapore’s first Underground Ammunition Vibration 14:305-320. Walker, R.E., Albritton, G.E. and Kennedy, T.E. Facility. Chee Hiong has also pioneered the development of protective 1966. The Elastic Response of Buried Cylinders technology within the Ministry of Defence. Under his leadership, the Leighton, T.G., Richards, S.D. and White, in Sand. Technical Report, Waterways protective technology centres in the National University of Singapore (NUS) P.R. 2004. Trapped Within a ‘Wall of Sound’: Experiment Station. and NTU were formed. Chee Hiong represents DSTA in the North Atlantic A Possible Mechanism for the Bubble Nets of Treaty Organisation Ammunition Safety Group (AC/326) Subgroup 5 Humpback Whales. Acoustics Bulletin, 29(1): Würsig, B., Greene, Jr. C.R. and Jefferson, T.A. Underground Ammunition Storage Custodian Group. He is also a Member of 24-29. 2000. Development of an Air Bubble Curtain the NUS Civil Engineering Department Consultative Committee. Chee Hiong to Reduce Underwater Noise of Percussive obtained a Master of Science (Civil Engineering) degree from NUS in 1986 Leong, E.C., Anand, S., Cheong, H.K. and Piling. Marine Environmental Research and a Master of Business Administration (Accountancy) degree from NTU Lim C.H. 2003. Response of Buried Structures 49:79-93. in 1993. Subjected to Below Ground Explosion. Paper presented at the 11th International Symposium on Interaction of the Effects 119 of Munitions with Structures, Germany, 5-9 May. Designing ABSTRACT In today’s network-centric battlefield, self-forming or self- Tactical Networks – healing networks play an important role in the operation of mobile forces that are deployed quickly to meet tactical Perspectives from a Practitioner demands. However, tactical network environments also pose significant challenges as a result of unreliable connectivity, limited bandwidth and latency issues. Research has focused on aspects including data link layers, routing and transport protocol as well as cross-layer optimisation. Many armies have large numbers of Very High Frequency and Ultra High Frequency narrowband tactical radios. The performance of these legacy radios can be optimised to support battlefield digitisation. This article suggests some guiding principles and key considerations in the design and implementation of tactical legacy radio networks for the Singapore Armed Forces.

Lai Ying Cheung Chia Wan Yin Designing Tactical Networks – 122 Perspectives from a Practitioner

for the design and configuration of tactical INTRODUCTION networks. The article also discusses possible challenges of adaptation for application Tactical networking facilitates information implementation, as well as the proposed LOS B LOS sharing and data exchange among military implementation of the various best practices A C tactical force units to enhance operational for tactical C2 systems. Non-LOS processes and situational awareness. It enables command and control (C2) capabilities for network-centric warfare TACTICAL OPERATING (Alberts, Garstka and Stein, 1999) as well as ENVIRONMENT A sends data to B C sends data to B network-based defence (Lundqvist, 2000). B The variations in tactical communication There is a need to understand the tactical network capabilities are dependent on operating environment in order to A C factors such as operational mobility, terrain, appreciate its characteristics in deploying communication media resources and network robust communication networks. Current Data collision occurs characteristics. To facilitate interaction and tactical communication networks support interoperability at different tactical data VHF/UHF line-of-sight (LOS) waveforms as Figure 1. A hidden node scenario networking levels, suitable adaptation at well as communication media e.g. Local Area Networks and Wide Area Networks. The key application level and network configurations exchange and reduce overheads. Messages Hidden Nodes challenge lies in addressing typical tactical need to be considered for deployment of large data sizes must not be sent to avoid communication network constraints such depending on the operational environment overloading the network and increasing The hidden node phenomenon is usually as limited bandwidth and latency, narrow and resources available. network latency. The number of nodes caused by the absence of the LOS between effective communication range, intermittent per network configuration should also be some networked nodes, which may be due communication links, and potential hidden Mobile ad hoc networking is seen as a assessed, as more nodes will increase the to mobility, terrain masking or weak signal nodes in certain operational terrains. In key enabler in realising the tenets of network data traffic and latency. strength. Figure 1 illustrates a scenario where addition, the mobility of tactical nodes network-centric warfare at the tactical networked node B is within LOS of nodes A causes rapid changes in network topology networking level. While current research and Intermittent Communication and C, whereas node A is out of LOS of node with the nodes leaving and joining the investigative work focus primarily on routing C. When node C transmits data, node B will network on an ad hoc basis. Links and transport protocol as well as wideband be able to relay this information via carrier radio waveforms, there is still scope for a sensing while it withholds data transmission A tactical network is usually required to Due to the high mobility of tactical nodes, mobile ad hoc network over a legacy radio from itself to avoid data collision. However, interact with a static strategic network which communication links may be broken due to network. This is because many armed forces as node A is out of LOS of node C, it is unable has a larger bandwidth and more reliable terrain masking. Thus, networked nodes may still have a sizeable number of narrowband to perform carrier sensing for any data communication links. Data exchange across appear online and offline intermittently, radios operating at Very High Frequency transmission from node C. Thus, data collision different network levels is required to support indicating that the communication links are (VHF) or Ultra High Frequency (UHF). may occur if node A has data to send. tasking to subordinate tactical-level force unstable and unreliable. The tactical communication network can units, and to allow reporting on-the-ground This issue needs to be addressed at the For LOS radio networks, there are several be designed to achieve effective data battlefield status to higher command. routing protocol and application layers. variants of Medium Access Control (MAC) communication and exchange. There is a Routing requires mechanisms such as ‘store- schemes such as Carrier Sense Multiplexing need for a structured approach that comes Limited Bandwidth and Latency and-forward’ to enable data to be stored Access (CSMA) and Time Division with a clear understanding of the operating at one of the networked nodes within Multiplexing Access (TDMA). The selection environment characteristics, communication The bandwidth of a tactical communication proximity of the destination node. When of the appropriate MAC scheme is critical media limitations and application data network is dependent on what the the destination node is back online, data will to ensure a successful adaptation to the exchange profiles. These form the basic communication media can support. If the be forwarded to it. This ‘store-and-forward’ operating environment. tactical network design considerations to most stringent communication media is feature saves network bandwidth as the define the relevant network parameters and adopted, there is a need to ensure that the sender node will not have to re-send the data CSMA is a distributed random access protocol values. The tactical network design is then network configuration is able to support to the destination node. At the application that is based on random carrier sensing before verified through simulation. VHF/UHF radios with low data rate. It is end, data buffering needs to be managed the initiation of any data transmission. It is 123 important to analyse the user’s data sending to avoid buffer overflow, while the data size also sensitive to the hidden node The main objective of this article is to requirements to address the problem of transmitted should be small to ensure better phenomenon. Literature reveals several present a practitioner’s perspective on the limited bandwidth. Pre-formatted data success of receipt. variants of CSMA implementation that guiding principles and key considerations messages should be used to streamline data Designing Tactical Networks – Perspectives from a Practitioner 124

attempt to overcome the hidden node SMEs possess a good understanding of the Payload problem. Collision avoidance mechanisms operating environment characteristics, radio Data Size Intended Recipients Reliability Frequency Type such as using the random carrier sensing limitations and application data exchange timer before data transmission and capturing requirements. Position 30 bytes All peers (Broadcast) No Periodic uni-directional communication links help to Location mitigate and prevent data collision (Chau, The understanding of expected payload Short Designated (Unicast, Depends on 150 bytes Ad hoc Chen and Liew, 2009). characteristics, types of applications and Messaging Multicast) context user-node mobility will shape network design Status Designated (Unicast, TDMA is based on the allocation of time slots and implementation at the application level. 100 bytes Yes Periodic Reports Multicast) for each networked node to perform data The proposed process shown in Figure 2 aids Designated (Unicast, transmission. If time synchronisation can be in analysing user requirements, determining Orders 250 bytes Yes Periodic achieved among the networked nodes, the the network design parameters and verifying Multicast) possibility of data collision is eliminated. the network design through simulation. Table 1. Application data payload classification (Source: Adapted from Nakamura, 2008) Upon achieving satisfactory simulation results, validation is conducted through field USER REQUIREMENTS and data sending frequency. Examples of trials and exercises in an actual operating also be derived to determine the average broad categories of payload are illustrated in ANALYSIS environment and over a representative time taken to deliver a message across the Table 1. set-up of the network structure. network, thereby providing estimates of the Besides understanding the intended expected network latency. Unsatisfactory operating environment, it is critical to analyse Key performance indicators are defined message completion timings could be due High Mobility of Tactical Forces user requirements for tactical communication for the assessment and verification of the to a number of reasons including poor networks. A structured approach is imperative simulation results. The average message network connectivity that requires data to be The high mobility of tactical forces results for various subject matter experts (SME) in completion rate will determine the reliability re-sent, or transmitting a sizeable payload in rapid changes to the network topology. communications and C2 to jointly analyse of the network in delivering the required over a network with limited bandwidth. If Network topology typically changes when the requirements as well as define the payload within the defined operational the simulation results are unsatisfactory, the networked nodes appear online and offline network and application parameters. These scenario. The message completion time can network parameters or scenario parameters intermittently. Depending on physical can be refined for simulation re-runs. location, the nodes may appear offline if they go beyond the effective communication Operating Environment range. User Requirements Tactical Command and Control Characteristics Applications The intermittent connectivity may occur C2 applications mainly facilitate situational when one node disconnects from its Operational Scenario awareness and the effective execution of current network and connects to another (Define payload, nodes mobility battle operations. Typical C2 application neighbouring network. Self-forming and connectivity) requirements are used to monitor the networks in such situations are required locations of friendly forces and adversaries, to maintain network connectivity. At the inter-network level, sent data are routed to Network Design as well as carry out navigation and battle Parameters management functions such as short the destination node even if the networked Refine messaging, status reporting and the issue/ node switches to connect to a neighbouring payload, execution of orders. network. Refine nodes Verification through Network mobility and Simulation Parameters connectivity Application Data Payload The defined network design and configuration, along with the data sending requirements and data dissemination It is necessary to manage and regulate implementation, can be verified using the data flow to allow critical information Simulation Results network modelling and simulation tools. exchange in an environment with limited Communication media models, operating bandwidth. Receipt acknowledgements for templates and data sending profiles can be data delivery must be used with care to avoid 125 set up to verify network design principles and overloading the network. The application Field Trials theories, as well as to assess the data traffic data payload can be categorised in terms loading effects and effective bandwidth of data size, number of intended recipients, utilisation. Simulation is necessary for large- requirement for receipt acknowledgement Figure 2. Proposed network design verification process scale network design and validation as the Designing Tactical Networks – Perspectives from a Practitioner 126

actual network set-up in the operating within its allocated time slot, each node is environment is impractical and resource- given an equal opportunity to send its data. intensive. TDMA is also suited for periodic data sending such as Blue Force Tracking i.e. periodic NETWORK DESIGN updates of peer node locations. However, TDMA requires control overheads such as FACTORS time synchronisation and keep-alive control messages to maintain the network. To achieve an effective network design, it is essential to study the inherent network Time Slot Allocation characteristics to identify and understand the boundaries of tactical communication In TDMA, the channel is broken down into networks. These design factors affect the time slots of equal duration and these are intra-network (single network) functionality, allocated to nodes for data transmission. Such Figure 4. Process of data transmission and receipt efficiency and performance of the network. slots can be pre-assigned prior to a mission, or Analysis of user requirements is required simplified breakdown of a time slot structure. To transmit data effectively, the design of the dynamically allocated as and when required. to obtain the intended deployment, data PTT delays that are dependent on the radio time slot allocation must take into account To perform the latter, it is essential to exchange pattern and application usage. hardware can affect the performance of the typical size of the data sent. For larger develop flexibility and operator competence. This analysis enables better evaluation of the network significantly (Navalekar et al., data sizes, time slots should be designed The number of slots in the TDMA system design choices and trade-offs for both tactical 2008). adequately to avoid the need to perform data will be influenced by the number of users network deployment and the application segmentation, as this would require multiple in the radio net. Hence, a balance between implementation. Figure 4 illustrates the process of data cycles to complete the data transmission. For efficiency and latency has to be maintained. transmission and receipt with a breakdown small data sizes, it is undesirable to allocate Longer slots increase the proportion of data of the relevant time-out expected. At the large time slots as the bandwidth would be Medium Access Control relative to overheads, while shorter slots can transmitter’s end, the Start-PTT latency is wasted. Schemes reduce latency between transmissions. required to ramp up the power amplifier before any data transmission can occur. The legacy VHF/UHF remains the most widely Therefore, it is critical to determine the Maximum Transmission Unit This is followed by the End-PTT time that used tactical communication media in the optimal length of the time slots to be indicates the end of transmission. Finally, Maximum Transmission Unit (MTU) refers battlefield. The choice of MAC schemes over allocated to each node. TDMA slot sizes are the transmission overhead time is required to the data payload that can be sent within the tactical communication media is largely determined by packet size and the amount to complete the sending of residual data. At an allocated time slot. This typically includes dependent on the operating environment of propagation delay expected in the the receiver’s end, pre-transmission and post- application data and header overheads of and radios used. CSMA is prone to the hidden system. A guard time is built into each slot transmission delays are taken into account in the preceding network layer information (see node phenomenon leading to the occurrence to ensure that transmissions do not overlap order to receive the data completely. Figure 5). of network instability at high network load and is guided typically by the longest possible (Braten, Voldhaug and Ovsthus, 2008). propagation delay. Theoretically, the performance of tactical However, CSMA has a lower control overhead Number of Nodes per Network network access is significantly influenced as compared to other MAC protocols and is The Push-To-Talk (PTT) time of a radio by the performance of the radio. Receive- well suited for quick bursts of very small data indicates the start and end of data Network size is determined primarily by to-Transmit (Rx-Tx) Turnaround time has a sizes. transmission. It is also known as the preamble the task group of nodes that communicate considerable impact on throughput, even time for the transmitter and receiver. The with one another frequently. However, in a for slotted configurations. Rx-Tx Turnaround TDMA is able to overcome the hidden node other time variable affecting effective data network environment of limited bandwidth, time refers to the time taken for a radio to phenomenon that is observed in some transmission involves setting a reasonable the network design will also need to take transit from receiver mode (Rx) to transmitter operating environments. By allowing each time to allow certain data sizes to be into account the expected update time mode (Tx). networked node to take turns and transmit sent within a time cycle. Figure 3 shows a cycle, bandwidth sharing among the nodes,

Start data End data transmission transmission PTT Start Data Transmission PTT Release 127

Start PTT Data Transmission End PTT Transport Comms Protocol Application Data Header Header

Figure 3. Simplified breakdown of a time slot structure Figure 5. Simplified illustration of MTU Designing Tactical Networks – Perspectives from a Practitioner 128

and data sending requirements. These update rate and network efficiency. Periodic dependency factors will affect the sending of network control messages should effectiveness of the network design and also be optimised for network efficiency by thereby influence the number of nodes per combining multiple control messages in a network. single transmission, or by synchronising with the rate at which application data is sent. A short update time cycle for better turnaround time cannot be achieved if the Network Throughput number of nodes per network increases. The available bandwidth and time slot In our work with the Singapore Armed allocation per node are fixed properties of a Forces (SAF), we observed that network communication network. With an increased throughput not only is influenced by the data number of nodes per network, the update rate of the communication media, it is also cycle time would also increase. If the design dependent on other factors such as time goal is to maintain the update cycle time, slot allocation and transmission overheads. there is a need to re-group the nodes into Doubling the data rate of communication multiple networks. media does not double the network throughput (Wang et al., 2007). The non-linear Alternatively, the data sending capacity relationship between network throughput Figure 6. Causal factors influencing network design for each node can be decreased to reduce and bandwidth is more significant for small the time slot allocation per node. However, data sizes than for large ones. The variable this method may not always be feasible network throughput is also due to unreliable which may take the form of ensuring that account bandwidth availability and user as it is subjected to the user’s data sending or intermittent network connectivity of network properties are suitably defined and requirements analysis, suitable adaptation requirements and the fixed portion of the mobile nodes which require multi-hop configured. In addition, adaptation mechanisms may have to be incorporated in time slot. routing. mechanisms are required to manage the data link routing protocol (middleware) the network load and utilisation during or in the application layer to ensure network Data Update Cycle Time Factors Influencing Network application. performance and efficiency. Design There is a common misconception that radios Communication Network There are several adaptation mechanisms with a higher data rate will improve the that should be considered when working Network design is a trade-off between radio Characteristics data cycle time. If the time slot allocation with a limited-bandwidth network: and network effects. Available network and number of nodes remain constant, the The first principle of network configuration design choices must be analysed to achieve the • Reduce transmission overheads. Data update cycle time will also remain constant. is to understand the network characteristics desired operational network effects. Figure 6 packets will ride on application data payload, However, a higher data rate allows more and identify their boundaries. Network provides a summary of the casual factors of while multiple small data packets are data to be sent within the time slot and characteristics include communication network design and their relationships. consolidated in a single data transmission. this is useful for large data transmissions. link reliability, bandwidth availability, Other network design considerations that network mobility, network access type and • Limit application data payloads. can increase network efficiency include GUIDING PRINCIPLES FOR supportability. If the network characteristics Methods such as data compression and pre- minimising transmission overheads, designing NETWORK DESIGN AND are clearly understood, the appropriate formatting application messages can reduce small application data sizes, and defining network properties such as the network CONFIGURATION application data payloads to levels that are the optimal periodic sending of control hierarchy level, expected latency and the supportable by available bandwidth. messages. The proposed step-by-step engineering gateway across transit networks can be guiding principles involve requirement configured effectively. • Adopt effective means of Large transmission overheads have a analysis with a clear understanding of users’ disseminating data. Methods include significant impact on update rate regardless needs, the operating environment and Data Bandwidth Availability broadcast messaging to send periodic location of the radio data rate. This is more evident network design factors. The operation of information, multicast messaging for multiple for small data sizes in view of the higher tactical communication networks requires The key element of engineering the recipients, as well as publish and subscribe overhead percentage compared to large data 129 the consideration of suitable adaptation network design configuration is to for quick dissemination of information across sizes. The application data sizes are designed mechanisms for designing both intra- determine the effective data bandwidth. transitional networks. to be small for the data link protocol to networks and inter-networks (across multiple The cost effectiveness of network routing is consolidate multiple small data packets in a networks). A holistic approach in network dependent on the efficient utilisation of • Prioritise data messages to minimise single transmission, thereby improving data design is required at the various layers, available data bandwidth. Taking into latency for urgent critical data. Designing Tactical Networks – Perspectives from a Practitioner 130

• Design application payload with data of fixed PTT time with tolerance. This leaves resolution differentiation for discarding less than one second to send useful data Communication Backbone Backbone detailed data when routing from high to low across the network. With a network of 10 Network A Comms Layer bandwidth networks. This mechanism allows nodes, each time cycle will be 20 seconds. This only critical data to be sent, and prevents a means that a networked node will be able to Access bottleneck from building up when routing send data every 20 seconds and the update Network B1 Network B2 Comms Layer through low-bandwidth networks, especially cycle time is 20 seconds. for potentially large data sizes such as images and file attachments. In addition, the optimal time slot configuration must include a buffer for Network Network Network Network Network Peer-to-Peer Comms Layer Data Sending Requirements network control messages, or have a provision C1 C2 C3 C4 C5 for gateway nodes to perform intra-network Bandwidth of Network A > Bandwidth of Network B Bandwidth of Network C To ensure success and effectiveness in or inter-network routing. network design, data sending patterns and requirements within and across networks Number of Nodes per Network Figure 7. Hierarchical architecture of a communication network have to be analysed. Data sending patterns Configuration can be inspected through the following should take into account the capabilities of Resolution of Critical Data ways: While sending data between nodes each communication medium to design the configured within the same network is overarching network architecture and deploy The size of data sent over communication • Identify application and control data that expected to be faster with inter-network the appropriate communication media at channels with limited bandwidth should must be sent periodically routing removed, an excessive number of the different network architecture levels. be regulated. Techniques to send data configured nodes in a network will affect with differentiated resolution need to be • Identify application data type and size its overall latency and responsiveness. The hierarchical communication network implemented to ensure that only critical that must be sent within and across There is a need to strike a balance among architecture illustrated in Figure 7 information is sent to the recipients with networks the dependency factors of expected update shows how a battlefield communication access to limited bandwidth. Detailed cycle time, bandwidth sharing among infrastructure is essential to provide routing information can be omitted to ensure that • Assess expected data routing flow, networked nodes, and data sending support across all network layers. Higher only critical information arrives at the including the identification of nodes that requirements. These dependency factors level networks are equipped with higher destination in a timely manner. require communication with one another, affect the effectiveness of network design. bandwidth communication media, which their frequency, and the number of inter- facilitates routing across peer networks Data Structure Format networking routing from sender to recipient In order to achieve an optimal update cycle that are linked to access networks. The nodes time for a better turnaround time, nodes battlefield communication infrastructure Operational data that is required for will need to be re-grouped into multiple will also facilitate network management transmission over tactical networks should After the data sending patterns are assessed networks. Data sending capacity for each at the higher hierarchical level to facilitate be structured as a formatted message and determined, the effective MTU can node will need to be assessed for the smooth information flow as well as network to minimise overheads. Data should be be computed based on the data sending opportunity to reduce the time slot allocation monitoring and control. represented in a numerical form instead of frequency and typical data payload required. per node. However, this method may not the actual value or full description i.e. bit- always be feasible as it is subjected to the oriented messages. It is not recommended to Time Slot Allocation user’s data sending requirements and the Adaptation at Application Level send large application data and files through fixed portion of the time slot. Configuration There are several application adaptation tactical networks. mechanisms and data dissemination To achieve effective data transmission, time Infrastructure Support for techniques that should be considered for Data Sending Frequency slot allocation must be optimised for the Battlefield Communication implementation at the application level to required data size and the fixed PTT time. manage network load and utilisation: The frequency of sending application data Time slot optimisation may be based on size The basic principle for establishing the and system management messages needs to of the data sent frequently, or the communication infrastructure is to leverage Data Compression be synchronised and controlled to regulate consolidated size of the data sent current network resources and capabilities to the overall data traffic over the limited- periodically. 131 achieve the communication network effect. Data compression techniques should be bandwidth network. A tactical communication network typically applied to further reduce the data size to be For example, a time slot allocation of two comprises heterogeneous communication sent over the tactical networks. seconds may include more than one second media. The network design configuration Designing Tactical Networks – Perspectives from a Practitioner 132

Acknowledgement and Priority Managing Classified http://ieeexplore.ieee.org/stamp/ http://ieeexplore.ieee.org/iel5/4723889 Information Flow stamp.jsp?tp=&arnumber=4753415 /4753027/04753150.pdf?arnumber=4753150 The operational data and supportability (accessed 16 August 2010) (accessed 16 August 2010) of data loading in the network should be To manage classified information flow across balanced. While it may be good to send Chau, C., Chen, M. and Liew, S.C. 2009. Capacity Navalekar, A.C., Kolanjery, J., Michalson, networks, it is critical to assign the appropriate application data at high priority together of Large-Scale CSMA Wireless Networks. W.R. and Matthews, J. 2008. Effects of security classifications to the application data with the delivery and receipt Paper presented at the Annual International Push-To-Talk (PTT) Delays on CSMA based and the network. acknowledgement, priority has to be Conference on Mobile Computing and Capacity Limited Land Mobile Radio (LMR) assigned to the data due to the limited Networking (MobiCom), Beijing, China, 20-25 Networks. http://ieeexplore.ieee.org/stamp/ bandwidth available. It is essential to exercise CONCLUSION September. stamp.jsp?tp=&arnumber=4556205 (accessed prudence in defining mission-critical data 6 August 2010) that require urgent and immediate sending, To ensure effectiveness of the network in Lundqvist, A. 2000. NetDefence: The Current and receipt acknowledgement. providing a reliable communication link in Revolution in Military Affairs. Military Wang, H., Crilly, B., Zhao, W., Autry, C. and the tactical land environment, several factors Technology No. 12, December 2000:72-73. Swank, S. 2007. Implementing Mobile Ad Hoc If the appropriate data dissemination for network design configuration have to Networking (MANET) Over Legacy Tactical mechanisms are used, bandwidth will be be considered. The proposed principles and Nakamura, M. 2008. Analysis of Alternatives Radio Links. Paper presented at Military utilised more effectively in data traffic design considerations provide guidance For Tactical Digital Battlefield Operation Communications Conference, 29-31 October. loading. The following data dissemination in evaluating available design choices and Over Low-Bandwidth Legacy Radios. mechanisms are recommended in view of analysing effective trade-offs for tactical data sending requirements. network deployment and application implementation. To achieve effective BIOGRAPHY Broadcast Messaging data communication and exchange over a networked system, C2 applications must take Broadcast messaging is an effective way into account the communication networking Lai Ying Cheung is Head Engineering (Communications). Besides leading of sharing data with peers within the same environment, and have the ability to adapt technical development in the Communications area, he also steers the build- network. By broadcasting one’s location and and factor in physical communication media up of future communications capabilities and harmonises organisational sending situation reports, all the networked constraints and network limitations. The engineering processes, methods and tools. He was involved in numerous nodes within the same network will receive ideal solution is to combine the designs communications related projects that have contributed to the build-up of the data. Data efficiency is achieved without of communication networks and C2 the Singapore Armed Forces (SAF) communications and networking having to send multiple copies of the same applications. capabilities. He was part of the Integrated Communication System Programme data to everyone in the same network. team which won the Defence Technology Prize Team Award in 1992. In the future, technology advances may He obtained a Master of Science (Electrical Engineering) degree from the Multicast Messaging lead to the availability of higher capability National University of Singapore (NUS) in 1990 and a Master of Business or hybrid communication media which will Administration (Management of Information Technology) degree from Data dissemination methods using multicast have to be supported. The guiding principles Nanyang Technological University in 1997. send a single copy to a recognised group and key design considerations should remain address. Hence, networked nodes that relevant but adaptation and configuration belong to the group address will receive the techniques may need to be reassessed for data sent. relevance. Chia Wan Yin is a Principal Engineer (Networked Systems). She is involved in systems integrations and data link related works, ensuring end-to-end Publish and Subscribe REFERENCES networked systems interoperability. She also manages the acquisition and implementation of Command and Control systems and data link The Publish and Subscribe mechanism allows Alberts, D.S., Garstka, J.J. and Stein F.P. 1999. communication protocol. Wan Yin is one of the recipients of the Defence users to subscribe to a specific topic or area of Network Centric Warfare: Developing and Technology Prize Team (Engineering) Award in 2006 for the design and interest. This offers an effective method for Leveraging Information Superiority. CCRP development of compact marine craft for the SAF. She obtained a Master of the quick dissemination of information to the Publication Series, 2nd ed (Revised). Science (Defence Technology and Systems) degree from NUS in 2010 and a relevant parties. Master of Science (Computer Science) degree from the Naval Postgraduate Braten, L.E., Voldhaug, J.E. and Ovsthus, K. School, US in 2010. 133 2008. Medium Access for a Military Narrow band Wireless Ad-hoc Network; Requirements and Initial Approaches. ABSTRACT Using Analytic Hierarchy The Analytic Hierarchy Process (AHP) is a structured decision making support tool that has been used by the Ministry of Process with Operations Defence and DSTA for evaluating weapon systems and platforms. Since 2005, AHP has also been adopted in the evaluation of Analysis in Project Evaluation several large-scale non-defence government tenders.

As the Singapore Armed Forces transforms into a Third Generation fighting force, new technologies and concepts of operations have evolved, resulting in military systems becoming highly interconnected and interdependent. Multiple and complex interactions among systems that are part of a larger system are expected during missions. Systems evaluation which requires assessing the military worth of a system in an operational scenario has thus become more challenging. Methodologies such as operations analysis and simulation techniques have been adopted to complement AHP in the evaluation process.

This article describes the successful adoption of AHP for the evaluation of several large-scale government tenders as well as the enhanced AHP methodology developed within the defence community for evaluating complex systems.

Kwok Yoong Fui Lim Hang Sheng Using Analytic Hierarchy Process with Operations 136 Analysis in Project Evaluation

Kiang announced that AHP would be used Weights denoting the relative level of Definition Intensity INTRODUCTION to evaluate the Marina Bay Integrated Resort importance are assigned to each criterion project. DSTA was engaged as a consultant using pairwise comparison – a method which Equal importance / quality 1 “Singapore’s evaluation was widely praised to the Singapore Tourism Board (STB) to compares two criteria that are on the same as being thorough, in-depth and provide advice on using AHP for the project level to determine their relative importance. Moderately more important 3 comprehensive, and the competition was evaluation. The comparison is done based on a simple / better hard fought.” ratio scale of one to nine as defined in Table Following the successful application of AHP 1. For example, when Capability is compared ‘Unique F-15SG is Perfect for Strongly more important 5 to the Marina Bay Integrated Resort project, with Availability, and Capability is assessed Singapore’s Fighter Project’ / better other government agencies in Singapore have to be moderately more important than Flight Daily, 21 Feb 2006 also consulted DSTA on the use of AHP. DSTA Availability, a weight of ‘3’ will be assigned Very strongly more 7 was engaged for several national projects to the former. important / better As the defence procurement agency for including the Singapore Sports Hub and the the Ministry of Defence (MINDEF) and the Changi Motorsports Hub projects by the The comparison process is used to assess Extremely more important 9 Singapore Armed Forces (SAF), DSTA employs Singapore Sports Council (SSC), the National all levels of criteria and the outputs are / better an objective and transparent procurement Broadband Network project by the Infocomm synthesised using eigenvectors to determine process where acquisition decisions are Intermediate values 2, 4, 6, 8 Development Authority of Singapore, as their respective weights in the hierarchy. For the result of comprehensive and rigorous well as the upgrading of the Changi Airport each criterion, the alternatives or choices technical evaluations (Soh, 2008). Table 1. Saaty’s scale for pairwise comparison Terminal 1 by the Civil Aviation Authority available are compared with one another of Singapore. The AHP technique proved to determine the overall score of each The defence procurement system has by a cost-benefit evaluation when the to be versatile and flexible with successful alternative. The alternative with the highest enabled DSTA to secure the best available price proposals are released. Evaluating applications to projects which differed score is the preferred solution. defence systems at the most competitive the benefits of an alternative without price significantly in scope and objectives. prices. The procurement system uses a information ensures a more objective and fair structured technique known as Analytic APPLICATION OF evaluation. Hierarchy Process (AHP) to support FEATURES OF THE ANALYTIC HIERARCHY acquisition decisions on major weapon ANALYTIC HIERARCHY PROCESS IN DSTA’S DSTA’s INVOLVEMENT IN systems and platforms. AHP provides a PROCESS PROCUREMENT SYSTEM GOVERNMENT PROJECTS rational framework for decision making by

breaking down the process into components AHP is a decision making support tool During the tender and evaluation process, Understanding the Project with respect to an overall goal. Alternative developed in the 1970s by Thomas Saaty, a Project Management Teams (PMT) employ solutions are then evaluated using a method mathematics lecturer from the University a two-envelope system (see Figure 2) in Prior to providing AHP consultancy services called “pairwise comparisons”. of Pittsburgh, US. The process requires the conjunction with AHP. This two-step process to other government agencies, DSTA’s establishment of a hierarchy of criteria consists of a technical evaluation of the competency in AHP lies mainly in the Up till 2005, MINDEF and DSTA were the only which is important to achieve the goal of the relative performance of the alternatives, evaluation of defence systems. The organisations in Singapore with extensive decision problem. A simple AHP hierarchy independent of price information, followed first-level criteria and some lower- experience in AHP. In December 2005, the used to evaluate the acquisition of a fixed- Minister for Trade and Industry Lim Hng wing aircraft is illustrated in Figure 1. Submission of AHP Goal, To buy the fixed-wing aircraft that meets Criteria and Weights by PMT Goal operational requirements DSTA Procurement releases first envelope Technical Evaluation by PMT using AHP – Technical Proposal 1st level Capability Availability Risk criteria Submission and Approval of AHP Benefit Scores

DSTA Procurement 2nd level Air-to-Air Air-to-Ground Survivability releases second envelope Cost-Benefit Analysis by PMT criteria – Price Proposal 137

Alternatives Aircraft A Aircraft B Aircraft C Cost-Benefit Results for Decision Making

Figure 1. An AHP hierarchy Figure 2. DSTA’s two-envelope tender evaluation system Using Analytic Hierarchy Process with Operations Analysis in Project Evaluation 138

level criteria used in these evaluations Given the different motivations of the have been well established and refined stakeholders, a strong facilitator was Goal: To select a competent consortium to finance, develop and operate the motorsports hub, over the years through applications in required to manage the evaluation process and contribute to the development of a sustainable motorsports ecosystem in Singapore multiple projects. Hence, the approach is in order to arrive at a set of criteria and relatively straightforward for DSTA PMTs. weights acceptable to all parties. The then Quality of concept Financial and business To tackle non-defence government projects, STB Chief Executive Lim Neo Chian, facilitated proposal (80%) sustainability (20%) the DSTA consultancy team had to understand by the DSTA consultancy team, led the the technical and business aspects of the evaluation process and played a pivotal

projects before working with the various role through his firm and objective Preferred venue Contributions to Contributions Attractiveness develop the centre to develop project teams to determine the project approach in managing the diverse for motorsports of commercial events (40%) of excellence for local mix and events objectives and criteria hierarchies. views of the stakeholders. Through motorsports training motorsports (10%) his persuasion and reasoning, the and education (20%) ecosystem Early involvement in these projects was stakeholders were able to arrive at an (10%) Distinctive Marquee and Financial critical as it allowed the DSTA consultancy outcome acceptable to all parties. and efficient sustainable sustenance team to understand the project, and map out design motorsports calendar a comprehensive evaluation approach that Criteria Used Maintenance identifies key requirements to be included in and facility Professional Light the tender documents. Project evaluation criteria differed Quality Quality and management masterplanning quantity of motorsports motorsports plans significantly among non-defence governemnt and branding international racing and industry advanced (R&D, logistics, Managing Diverse Stakeholders projects due to their varied nature, which can events driving manufacturing Variations to be illustrated using two examples. Maximise academy and engineering) legal terms MINDEF is typically the main stakeholder positive Quality and and experience to of defence projects. For projects involving Figure 3 shows the first-level criteria quantity Effectiveness High conditions motorsports of regional users from different Services, there is often of the Marina Bay Integrated Resort of local performance participants events motorsports motorsport- a consensus on the key project criteria and project and key stakeholders for each training / related Suitability of their relative weights since the overall criterion. The goal of the project Maximise Quality and education businesses commercial proposals and objective is to achieve robust systems was kept generic as “Selecting the proposal positive quantity of mix experience to commitment effectiveness at an affordable cost. that maximises economic benefits for local events spectators / Attractiveness Singapore” (STB, 2006). This goal was event organisers Exportability Effectiveness of year-round However, for commercial projects which supported by the first-level criteria of of national of regional / leisure, involve several stakeholders, deciding on the tourism appeal, architectural excellence, motorsports international entertainment / motorsports programmes lifestyle events weights for the criteria is challenging due development investment and strength of training / education and activities to different interests of the stakeholders. consortium. While architectural excellence proposals and commitment For example, the key stakeholders in the does not contribute directly to economic Marina Bay Integrated Resort project are gains and therefore should not have a higher STB, the Ministry of Finance (MOF) and the weight than development investment, it Figure 4. AHP Tree for the Changi Motorsports Hub project Urban Redevelopment Authority (URA) – can be argued that an iconic monument whose areas of interests are tourism appeal, could lead to higher tourism appeal and development investment and architectural investment. Thus, a significant weight of 30% was allocated to architectural concept and of concept proposal – which was given 80% of excellence respectively. design, which was only 10% lower than the the weight. The other 20% went to financial 40% weight allocated to tourism appeal. and business sustainability to ensure that the winning bid has sufficient finances for the project’s design, development and operations Selecting the proposal that maximises economic benefits for Singapore For the Changi Motorsports Hub project, SSC’s goal extended beyond building in the long term. Figure 4 illustrates the goal a world-class race track. It had a more of the Changi Motorsports Hub project, the encompassing objective to promote and first three levels of criteria and some of the Tourism appeal Architectural Development Strength of develop Singapore’s motorsports ecosystem, weights (SSC, 2010). and contribution concept and design investment consortium and which includes motorsports training, the 139 (40%) (30%) (20%) partners (10%) motorsports industry as well as commercial The derivation of the criteria for each project set-ups and events to augment the appeal of was not a straightforward matter as these Key STB / MOF URA MOF MOF / STB Stakeholders : racing activities. These objectives were then projects were diverse in nature. Each project grouped under the first-level criteria – quality team had to agree on the overall goal of the Figure 3. Goal and first-level criteria for the Marina Bay Integrated Resort project (Source: STB) Using Analytic Hierarchy Process with Operations Analysis in Project Evaluation 140

project before the first-level criteria could be in operational scenarios over time and (2003) is adopted i.e. “Operations Research replaced with tactical and mission analysis. established. This process to arrive at the final space are often too complex to be deduced is a scientific method of providing executive For example, tactical analysis can help to hierarchy could take several weeks of intense intuitively by human perception or a panel of departments with a quantitative basis for assess the effectiveness of a fixed-wing deliberation. experts. decisions regarding the operations under aircraft against another fixed-wing aircraft their control”. In other words, OA is a in a single engagement scenario. Mission STRENGTHS AND In addition, different stakeholders also systematic and iterative approach that uses analysis helps to assess the effectiveness LIMITATIONS have different ideas of utilising the systems analytical techniques to provide a measurable of combat air patrols to defend against capabilities to conduct a mission. While AHP result on an operational issue. Its purpose is to incoming airborne threats. has been employed in many acquisition facilitate decision making among operational A key strength of AHP is the ability to projects, MINDEF is continually seeking to commanders, key stakeholders and decision The key benefit of OA is its ability to consider incorporate a variety of tangible and enhance AHP evaluation approach so as makers. These analytical techniques include both engineering and operational factors intangible criteria into the same hierarchy to better determine the military worth of statistics, probability theory, queueing theory, in a dynamic scenario to determine the to allow comparison using the same ratio increasingly complex and interdependent game theory, optimisation and simulation. military worth of the various alternatives. It scale. AHP facilitates group dynamics and weapon systems and platforms operating can also incorporate different concepts of organises feelings, intuition and logic from as a System of Systems (SoS). The complex As OA is mathematical and computational in operations from various operational users different stakeholders using a structured interactions among these systems and the nature, it relies heavily on computer science and stakeholders. As a result, the force approach to enable objective decision multiple roles they play in numerous mission technologies to develop tools and models for multiplier effects of a candidate system that making. Having a structured approach also scenarios require more comprehensive analysis. With the advancement of computing can be integrated with a suite of networked minimises the unintended exclusion of any evaluation tools. Hence, there is a need to technologies, Modelling and Simulation has sensors and weapon systems were observed criterion that could be important to the develop new methodologies to complement evolved into a mature discipline with wide- when compared with stand-alone systems evaluation outcome. the traditional evaluation approach. This has ranging applications, and it has also become and platforms that offer limited or no led DSTA to leverage Operations Analysis an essential technology and tool used in OA. integration with the SAF networked SoS. AHP is also simple to use as government (OA) to offer a more encompassing approach More importantly, the military utility of agencies with no prior knowledge of AHP as part of an enhanced AHP framework to For evaluations that require the use of OA a candidate system applied to different were able to apply it to their projects after support tender evaluation. to complement AHP, the benefits assessment concepts of operations could be quantified, going through some basic training. The provides inputs to the main branches of providing greater clarity for different criteria used for these projects are generally the AHP tree as shown in Figure 5. For stakeholders. not mission-oriented and are relatively ENHANCED the Capability branch, traditional factors independent of one another, making AHP an METHODOLOGY USING such as air-to-air capability, air-to-ground OA also offers insights into potential weak ideal tool. The swiftness with which AHP has OPERATIONS ANALYSIS capability and survivability have been links in systems design, rules of engagement been learnt and applied in diverse projects, TOOLS as well as the project teams’ satisfaction with the outcomes, testify to the efficacy of AHP. OA, also known as Operations Research, As a result, MOF has decided to mandate has been associated with systems analysis, the use of AHP for the evaluation of all To buy the fixed-wing aircraft that meets systems engineering, management science Goal operational requirements government projects costing more than S$50 and cost-effectiveness analysis (The RAND million. This policy was included in MOF’s Corporation, 1968). Its origins can be traced revised Instruction Manual on Procurement to World War Two when UK and US scientists 1st level issued in June 2009. Capability Availability Risk applied OA techniques to search and destroy criteria enemy submarines, protect merchant ships In defence applications, evaluation using and minimise the loss of aircraft from 2nd level Inputs from Traditional Traditional AHP has become increasingly challenging bombing missions. Today, OA is applied in criteria Operations Analysis Factors Factors for complex acquisitions as AHP cannot many areas beyond military applications model the dynamic scenarios that are e.g. industrial engineering, supply chain prevalent in the SAF’s network-centric management, as well as business and Alternatives Aircraft A Aircraft B Aircraft C operations. This is because the factors defined financial management. in the traditional AHP approach are treated independently, and their interdependence OA focuses on the operational nature of Figure 5. Enhanced AHP hierarchy and interactions in a network-centric 141 the issues that are being studied. In this operation cannot be adequately accounted respect, the definition by Morse and Kimball for. The consequences of such interactions Using Analytic Hierarchy Process with Operations Analysis in Project Evaluation 142

and supportability issues. For example, an OA analysis in systems design can reveal that REFERENCES BIOGRAPHY leveraging a platform with high endurance may result in more on-board systems failures Morse, P.C. and Kimball G.E. 2003. Methods Kwok Yoong Fui is a Principal Engineer and Resident Lecturer (DSTA College). and poorer mission effectiveness. of Operations Research. Dover Publications, Inc. He oversees the DSTA Basic Systems Engineering and Management Course and conducts lectures in Systems Engineering, Life Cycle Management, DSTA has been using the enhanced AHP Project Risk Management and Analytic Hierarchy Process (AHP). He plays methodology to evaluate weapon systems Saaty, T.L. 1990. Decision Making for Leaders: a key role in advancing AHP and provides advice on the use of AHP for and platforms such as the F-15SG Fighter The Analytic Hierarchy Process for Decisions tender evaluations, as well as issues related to life cycle management and Aircraft, Apache Longbow Attack Helicopter in a Complex World. McGraw-Hill, New York. project risk management. Yoong Fui was instrumental in the development and the Formidable-class Stealth Frigate, hence of the Project Risk Management methodology. He obtained a Bachelor of validating the effectiveness of the enhanced Singapore Tourism Board. 2006. Marina Bay Engineering (Electrical Engineering) degree with Honours from the National AHP framework. Stakeholders and decision Integrated Resort Evaluation Model. University of Singapore (NUS) in 1988. makers were able to obtain a better understanding of the effects of an acquired Singapore Sports Council. 2010. Changi system in an operational context. Motorsports Hub Evaluation Model.

CONCLUSION Soh, K.P. 2008. Speech at the International Lim Hang Sheng is a Lead Analyst (DSTA Masterplanning and Systems Defence Procurement Conference, Architecting). He has worked on integrating systems on fighter platforms 18 February. and is currently conducting systems effectiveness studies. He has also served The use of AHP for evaluation has proved as Senior Manager (Organisational Change and Learning) in Corporate to be successful despite the varied nature The RAND Corporation. 1968. Systems Analysis Planning and Development. Hang Sheng is a member of the project team of government projects. The criteria used and Policy Planning: Applications in Defense. which attained the Defence Technology Prize Team (Engineering) Award for these projects are generally not mission- in 2001. He obtained a Master of Science (Electrical Engineering) degree oriented and are relatively independent of from NUS in 1999 and a Master of Business Administration (Management of one another, making AHP an ideal tool. For Technology) degree from Nanyang Technological University in 2002. Under systems which have criteria that are more the DSTA Postgraduate Scholarship, he graduated in 2007 with a Master of dynamic and interdependent, the use of Science (Defence Technology and Systems) degree from NUS and a Master OA to complement AHP produces a more of Science (Operations Research) degree from the Naval Postgraduate representative and accurate assessment of School, US. the systems being evaluated.

143 ACKNOWLEDGEMENTS

The Editorial Board, DSTA Horizons

Chairman Technical Editor

Teo Chin Hock Professor Bernard Tan Deputy Chief Executive (Strategic Development) Department of Physics, Faculty of Science National University of Singapore

Members Peer Reviewers Pang Chung Khiang Director (DSTA Masterplanning & Systems Teo Chin Hock Architecting)/Director (DSTA College) Deputy Chief Executive (Strategic Development)

Lee Chee Teng Pang Chung Khiang Director (Procurement) Director (DSTA Masterplanning & Systems Architecting)/Director (DSTA College) Chia Lip Teck Director (C2IT) Lim Chee Hiong Director (Building and Infrastructure) Vincent Lim Buan Sui Acting Director (Human Resource) Wee Kok Ling Director (Networked Systems) Chew Keng Cheow Deputy Director (Guided Weapons Chau Chee Chiang & Armament) Deputy Director (National Security) – (till 2 March 2011) Alex Lee Yeaw Lip Deputy Director (Systems Engineering) Chew Keng Cheow Deputy Director (Guided Weapons Seah Peng Hwee & Armament) Deputy Director (Sensing & Connectivity) Alex Lee Yeaw Lip Koh Tong Chia Deputy Director (Systems Engineering) Assistant Director (Building and Infrastructure) Seah Peng Hwee Quek Bee Tin Deputy Director (Sensing & Connectivity) Senior Manager (DSTA College) Leow Aik Siang Pearly Chua Siew Ting Assistant Director (Knowledge & DSTA Systems) Senior Manager (Corporate Communications) Dr Koh Hock Seng Too Meng Yuen Head Capability Development (Surface Senior Executive (DSTA College) Warships) Jolene Chng Wang Chin Ong Chin Ann Executive (Corporate Communications) Chief Information Officer Desiree Tan Ting Ting Pee Eng Yau Executive (Corporate Communications) Lead Analyst (DSTA Masterplanning & Systems Architecting) Articles may be reproduced only with prior permission obtained from DSTA, and due acknowledgement given to DSTA. Please send enquiries, feedback and suggestions to:

The Editorial Board, DSTA Horizons DSTA College 1 Depot Road Defence Technology Tower A Singapore 109679

Email: [email protected]

DISCLAIMER: The declarations, opinions, references, or the views expressed in this publication are those of the individual authors concerned and do not necessarily reflect those of the Defence Science and Technology Agency.

All information is correct as at time of publication.