Cyber security of intelligent buildings: a review
The concept of an intelligent building has existed for over 30 years and is seen as a solution to a number of economic and environmental problems. Intelligent buildings are often seen as a means of improving the energy efficiency of the built environment. However, they are increasingly being promoted as a means of improving the operational efficiency of buildings, providing enhanced building and personal security, and improving the availability of management information to support better and quicker decision making. The potential cyber security risks inherent in the adoption of intelligent buildings are rarely discussed.
Author(s): Hugh A. Boyes
The potential demand for more intelligent buildings is a response to societal pressures to reduce the impact of buildings on the natural environment, for example, through more efficient use of natural resources in the construction, maintenance and operation of buildings, and the ability to make efficient use of space across a building's lifecycle. A particular target for this increased efficiency is use of technology to reduce energy consumption. From a building owner or occupier perspective, while resource efficiency is important, the expectation is that an intelligent building will provide a degree of automation allowing flexible use of the accommodation and providing tools or applications which support the occupation and management of the space.
To deliver intelligent solutions which improve the efficiency of the building and meet the owner and user's aspirations regarding the use of the space, there has been pressure to integrate systems. There is also a trend to make increasing use of a wide range of building sensors to acquire occupancy and usage information. Associated with these developments is an increasing use of the Internet to exchange information with third parties and deliver externally hosted services. This integration and connectivity is not without risk, as it exposes the owners and occupants to a variety of cyber security threats.
As noted by Fisk , the intelligent building community might protest that there have been no significant successful reported attacks on building management systems (BMS) over the 40 years that the technology has been deployed. This is potentially a complacent view as a range of complex threats are rapidly developing from increasingly sophisticated threat actors.
The first part of this article examines what is meant by an intelligent building and how from a cyber security perspective the convergence of technologies and integration of systems creates vulnerabilities. It then reviews how cyber security affects the built environment, examining the nature of some of the threats and identifying some key management challenges. This article includes a case study examining these issues in the context of an international transport terminal. It then discusses some of the management challenges that need to be addressed to reduce cyber security threats to the built environment.
What is an intelligent building?
The concept of intelligent buildings is not new. Clements-Croome  notes that the term ‘intelligent building was first used in the United States in the early 1980s and that a definition given by the Intelligent Building Institution in Washington was: ‘An intelligent building is one which integrates various systems to effectively manage resources in a coordinated mode to maximise: technical performance; investment and operating cost savings; flexibility’. The innovative idea of a building operating with a converged technical infrastructure as a ‘computer integrated building’ was first proposed by DEGW/Teknibank , as illustrated in Fig 1. Putting their pioneering work into perspective, this proposal for the convergence of systems based on an IP-network infrastructure was published before the Ethernet (TCP/IP) was generally accepted as a universal protocol for communications, before the creation of the Internet as we now know it, and several years before any commercial trading took place on the Internet.
Fig 1: Infrastructure convergence – the integration pyramid 
Since this early work by DEGW and Teknibank, there have been a number of definitions of what is meant by the term intelligent building. For example, Ehrlich  defines it as ‘use of technology and process to create a building that is safer and more productive for its occupants and more operationally efficient for its owners’. This definition reflects the needs of owners and occupiers, but does not explicitly address the environment. Another definition is offered by Harrison , an intelligent building ‘provides a responsive, effective and supportive intelligent environment within which the organisation can achieve its business objectives’. Underpinning these definitions is the Integration Pyramid model and increasingly through IP convergence, the integration of the multi-function building and user systems with business systems.
The range of systems that may typically be integrated within an intelligent building is illustrated in Fig 2 . For a building to be intelligent it is not essential for all of these systems to be present, integrated or to share a common infrastructure. It is the systems integration and sharing of data to deliver additional functionality that creates an intelligent building.
Fig 2: Typical systems in an intelligent building
It is important to recognise that wireless technologies play an increasingly important part in the network and communications infrastructure. The term ‘wireless’ in Fig 2 is used as a generic term to cover communications and data links which do not require a physical connection, technologies currently employed include Wi-Fi, Bluetooth, ZigBee, radio, NFC and RFID. The inclusion of IP-based applications is relevant where they interact with building systems or sensors, for example, RFID for tracking location of material or assets. The business systems are relevant only to the extent that they are integrated with building systems, for example, the use of CRM systems to manage access control, or the use of ERP/MRP systems to support supply chain management such as consumables and utilities.
Intelligent building case study – airport terminal
A consequence of the continuing growth of air transport, for example, a 65% increase in global scheduled passenger volumes over the last decade , are significant economic and environmental pressures on the owners and operators of international air terminals. Airports need to handle these increasing passenger volumes safely and securely, while reducing the cost of ownership of the terminals and their environmental impact. The transport industry is addressing these problems through the use of innovative IT-enabled solutions, allowing them to achieve energy savings and to increase passenger capacity of terminals.
Modern airport terminals are now typically delivered with a converged IP-based network infrastructure , based on a common cabling system. This networking infrastructure is required to support a diverse range of operational, business and facilities management systems as illustrated in Fig 3. Economic pressures for operational efficiencies have created a demand for increased integration of applications, which allows streamlined handling of data and eliminates the need for data re-entry thus reducing risks of errors.
Fig 3: Terminal-related IT systems architecture 
By adopting common, compatible physical distribution and networking layers in the architecture, as shown in Fig 3, the systems delivering applications use industry standards protocols to communicate between any distributed elements and via the systems integration layer, with one another. For example, from a central control room, an operator may be able to control landside systems using dynamic signage to provide directions for passenger vehicle traffic. Using sensors in parking bays, the system may then automatically indicate to drivers the location of available parking spaces. If a customer has pre-booked parking, a space may be allocated and barrier access controlled using CCTV and ANPR (ANPR – Automatic Number Plate Recognition) systems. The security of the terminal and its adjacent highways and parking areas will typically be overseen from a security control room, using IP-based CCTV and security systems. From a security perspective ANPR may be used for the identification and monitoring of suspicious vehicles in the public highway and terminal approaches. From a safety perspective tampering with or failure of signage on vehicle routes could cause vehicle accidents and injuries to pedestrians.
Within a terminal, sharing of passenger data between systems may allow Internet check-in, make baggage handling more efficient, and allow the customs and border security staff to monitor arrivals and processing of passengers. In addition to the use of IT systems to provide travellers and visitors to the terminal with up-to-date arrivals and departures information, customers increasingly expect to have Internet access with public access in terminals and Wi-Fi readily available both land and air side.
In the airport terminal, a common infrastructure will typically be used to support a range of network-based services , including: CCTV, access control systems, fire alarms, badging systems, computer aided dispatch, incident management and visa and passport control. Given an operational need to flexibly manage terminal accommodation, particularly in public areas, the building management systems, internal signage systems and access control systems are all likely to use this common infrastructure. Through appropriate systems integration, this allows passenger flow in corridors and stairways to be switched between arrivals and departure areas to suit the operational needs of the airport. From an environmental perspective this integration may also allow the heating, lighting and air conditioning of areas to be reduced when they are not in use by passengers.
This integrated infrastructure will also be used to support the operation of the terminal from a business perspective, for example, to provide HR services, staff email and internet access, asset management and so on. There may be some integration of business systems and security systems to enable flexible management and allocation of staff, based on their skills and security clearances.
This case study illustrates how an intelligent building may depend on a diverse range of systems, using common commercially available technologies. The integration and interaction of these systems increase the probability that the design will include significant and potentially unforeseen cyber security vulnerabilities.
Cyber security in the built environment
Defining cyber security
In the built environment, it is important to recognise that cyber security affects more than the IP-based networks in a building. An internationally agreed definition  which recognises this broader scope of cyber security is ‘the collection of tools, policies, security concepts, security safeguards, guidelines, risk management approaches, actions, training, best practices, assurance and technologies that can be used to protect the cyber environment and organisation and user's assets’. This definition refers to the ‘cyber environment’ (or cyberspace), which effectively comprises the interconnected networks of electronic, computer-based and wireless systems. The definition also refers to ‘organisation and user's assets’, which effectively includes all connected computing devices, personnel, infrastructure, applications, services, telecommunication systems and the totality of transmitted, processed and/or stored data and information in the cyber environment. In protecting the ‘cyber environment’ we also need to address the safety of those that could be affected by systems failure or malfunction.
Cyber security has developed from the discipline of information security, which was based on three core principles :
- Confidentiality – which encompasses privacy, and the control and authorisation of access to data or information, and any ability to process, modify or delete data or information;
- Integrity – which addresses the trustworthiness of the data or information storage, the authenticity to data and results, and the safe operation of electronic systems; and
- Availability – the availability of the systems and associated business functions when needed.
These principles are commonly referred to as the CIA triad. When considering the security of corporate IT systems, the order of triad is often seen as reflecting the priority of these attributes. This contrasts with the needs of control systems, for example, those used for building management and security where availability and then integrity typically have a higher priority than confidentiality. This combination of availability and integrity may be considered as resilience, that is, the ability to adapt and respond to disruption and maintain functionality.
Application in the built environment
In an intelligent building, the convergence of infrastructure technologies (e.g. networks and sensors), and systems integration can create significantly increased cyber security risks. Without careful implementation, the complexity inherent in this ‘system of systems’ may result in the application of the ‘Law of Unintended Consequences’ .
The complexity of applying cyber security to systems in the built environment inevitably becomes more complex as the degree of integration between systems increases. For example, sensors within a building may be used for multiple purposes. The output from a CCTV camera may be used by:
- an operator to monitor and observe individuals in the building;
- security systems to detect suspicious objects;
- alarm systems to detect movement in an area which is currently out of bounds or not in use; and
- building systems to save energy by restricting or shutting down environmental conditioning systems in unoccupied areas.
These applications will require feeds of digital data to systems with different security profiles and sensitivities, some of which may be located outside the terminal building. Failure to implement adequate cyber security measures could result in data loss, with images being viewed by unauthorised individuals, or even broadcast over the Internet. The exposure of the CCTV feeds may cause serious reputational damage to the terminal operator, particularly if individual passengers can be identified.
From a cyber security perspective, the greatest risk to an intelligent building is likely to be from the potential impact on the building control systems, in particular any loss of view or loss of control. Building management systems are typically designed to operate on their own dedicated network with security providing access control at operator workstations, but little or no security implemented between control computers and sensors or the systems being controlled. Unauthorised direct access to this network could allow a user to interfere with building systems, resulting in loss of control or view and causing significant safety or security issues, for example, an inability to shutdown mechanical equipment should a life threatening incident occur. Similarly, unauthorised modification of data could undermine systems integrity causing operator errors or loss of control. These situations could arise in an intelligent building if there is inadequate security between any networks used by office staff and/or the public, and the building management network.
Threats to intelligent buildings
Source and nature of threats
Cyber security threats to an intelligent building will emanate from four types of threat agent: malicious outsiders; malicious insiders; non-malicious insiders; and nature . The threats from malicious agents may be indiscriminate, such as distribution of malware or viruses, or targeted attacks attempting to compromise, disrupt or damage specific systems. Threats from nature relate to damage or interference to building systems, arising from solar, weather, animal or insect threat agents.
The malicious threats will typically originate from one of the following groups :
- Sole activists
- Activist groups
- Organised crime
- Proxy terror threat agents with nation state support
- Nation states
The above list is ordered to reflect the increasing capacity for damage and potential sophistication of threats originating from the threat group. The nature of malicious threats will vary depending on the motivation and objectives of a threat group. Their intent may be to cause commercial or reputational harm, to steal intellectual property or to simply cause disruption.
The airport terminal described earlier in this paper could be a target of malicious cyber security attacks by any of the above threat groups. Their intent may be to disable systems for terrorist or criminal purposes. For example, were a robbery to occur today, like the one which occurred in November 1983  – involving the Brink's-MAT warehouse at Heathrow Airport, London, it is likely that the perpetrators might have used cyber security attacks in their planning or execution of the robbery.
To identify and address potential cyber security vulnerabilities a risk management cycle  will typically be used as illustrated in Fig 4. The objective of using this cycle is to analyse potential threats, allowing development of proportionate safeguards to enable suitable mitigation to be put in place. When considering cyber security related risks affecting an intelligent building, the owner/operator can choose to avoid the risk, to reduce the risk, to share it or to retain it.
Fig 4: Typical risk management cycle
In an international airport terminal maintaining security of the air-side/ground-side interface is critical for effective immigration and customs control and to deter the air transport related terrorism. It is essential that availability and integrity of these systems are maintained. The design of new security systems (access control, CCTV and badging systems) are TCP/IP network based. If, for example, following an assessment of vulnerabilities, a terminal owner/operator may determine that the following cyber security vulnerabilities should to be addressed as they could lead to physical security breaches:
- the potential for loss of systems availability because of mains power failure;
- the use of wireless networking would be susceptible to jamming or interference thus affecting network availability; and
- the risk of unauthorised access and data corruption if the security systems infrastructure is not adequately protected where is passes through public areas. This may affect both confidentiality (access control) and integrity (of systems data).
In assessing risks associated with these vulnerabilities, the terminal operator can decide to avoid the risk through the following measures:
- to prevent power loss, supply these systems with electricity from an uninterruptible power supply supported by back-up generators;
- to only use cabled networks thereby eliminating the risks of jamming or RF interference; and
- to avoid unauthorised access in public areas, the cables may be routed via secure distribution ducts preventing physical access to the cable distribution system. Alternatively, the terminal owner/operator may choose to implement appropriate network security measures (e.g. firewalls, intrusion detection systems etc.) to maintain flexibility while reducing risk of attacks from outside the secure area.
Throughout a building's lifecycle those responsible for managing it will face a number of challenges which affect its cyber security. During the specification and design stages an important issue may be to protect investment in the intellectual property rights. This may be of particular significance where a design involves proprietary or innovative features or techniques.
The use of Building Information Modelling (BIM)  potentially introduces a number of cyber security risks, including:
- the potential for loss or unauthorised access to sensitive designs;
- management of data, including configuration management; and
- the exposure of site security vulnerabilities allowing vulnerability reconnaissance without need for physical access to the site or building.
The latter point may be a particular issue for sensitive buildings such as banks, prisons, military establishments and sites supporting the critical national infrastructure.
Building security has traditionally involved creating and maintaining a secure physical perimeter. This typically involves minimising and controlling the points of entry and egress. The interconnection of business and building systems with the Internet significantly increases the attack surface for any building, as a potential intruder now has both physical and logical avenues to attack. Internet connectivity also removes the need for an attacker to be near or inside a building or site. This challenge is particularly significant where a building's security and alarm systems operate over the building's IT infrastructure. A failure to adequately secure these systems could allow a cyber security attack to remotely disable or disrupt them.
It is essential to recognise that insiders, that is, those with authorised access to the building and its systems, can be a significant cyber security threat. An attack from a knowledgeable and malicious insider could be particularly devastating as the attacker may have a good understanding of which systems to attack or disable. However, negligent or careless insiders can also cause significant disruption or damage, for example, by plugging malware infected removable media into a building or business system. Failures by insiders to properly execute security procedures, to monitor alarms and system logs or correctly configure/maintain security systems and access control will significantly increase the security vulnerabilities.
With the increasing convergence of building and business systems, and the complexity associated with this systems integration, there is an increased need for greater collaboration between IT and facilities management teams. For example, these teams need to collaborate about the management of interfaces between their systems and the security measures in place to protect them. Without this collaboration there is a risk that the business systems and their interfaces may be used to mount an attack on the building systems or vice versa. Collaboration will require team members to have an appreciation of how different types of system work and may be protected. This is important as some techniques used to protect an administrative system may be inappropriate or not feasible on real-time control systems.
Another potentially significant management challenge is the on-going systems engineering and management throughout the building lifecycle. A significant difference between business and building systems is their typical lifetimes. A business system might typically be operational for 3 to 5 years, whereas a building system may have a lifetime of between 5 and 20 years . The implication of these differing lifetimes is that over a building's operational lifecycle there will be a number of changes as business systems become obsolete and are replaced. The impact of these changes will need to be assessed and given the increasing complexity of intelligent buildings this may require in depth systems engineering knowledge of all affected systems.
This paper has examined the development of intelligent buildings. Since the prescient prediction by DEGW and Teknibank in 1992 , their vision of an intelligent building is being realised through infrastructure convergence. Current economic and environmental pressures encourage development of intelligent buildings to address these societal needs. The current generation of intelligent buildings is potentially an important stepping stone towards innovations such as smart cities and the smart grid.
Unfortunately, the elements of a building which make them smart or intelligent also create significant vulnerabilities. While there are few reports of attacks on building systems , the increasing incidents involving malware such as Stuxnet , Flame  and Duqu  suggest that it is only a matter of time before building systems are affected. This risk is likely to be increased because of the existence of search tools like Shodan, and relative concentration of systems suppliers and commonality of components and sub-systems between a wide range of industrial control systems. For example, a component or sub-system used in a building may also be found in a manufacturing or distribution system.
Owners and operators of intelligent buildings face a number of significant management challenges if they are to protect the building adequately from a variety of cyber security threats. Possibly the greatest challenge will be to achieve a single team solution to maintenance and operation of both business and building systems. This situation will be further complicated if the current cultural, contractual and skills differences between these teams are not addressed. For example, it is common practice for organisations to outsource the majority of their facilities management activities, whereas the level of outsourcing within a corporate IT department can vary across a wide spectrum. In an intelligent building, a matrix of contracts may therefore exist with a number of contractors sharing responsibility for the overall cyber security of the building. This can significantly increase the complexity of maintaining adequate defences and responding to cyber security incidents.
The continuing rapid development of technology and solutions will pose another significant challenge for the building manager, for example, the use of ‘cloud services’. The airport terminal referred to in this article is a typical building where cloud services may be used for the business systems. The consequence of failure of these services could be disruption to the business operations, for example, checking-in of passengers. While disruptive, this is unlikely to affect the health or safety of terminal users. However, building systems suppliers  are developing solutions employing ‘cloud services’ for building management. If these building services delivered via ‘cloud services’ are critical to the safety and secure operation of the building then any interruption to these externally hosted services or connectivity to them could seriously disrupt the use of the building. From a resilience and cyber security perspective, availability and integrity of these ‘cloud services’ then becomes critical to the safety and security of terminal users.
The increasing use of BIM, a transformational initiative within the architecture, construction and engineering (ACE) industries has significant cyber security implications. As these industries move from their current use of BIM at Level 2 maturity (i.e. managed CAD in two-dimensional (2D) or 3D format) to Level 3 (iBIM – a managed 3D environment), cyber security risks will increase unless appropriate steps are taken. Increased sharing of electronic data between companies in the building design and construction supply chains has implications for both the security of the building models and because of the connectivity to the organisations involved. These problems do not cease on the handover of a building to the owner as BIM models are intended to be used by an owner and their facility manager to support the ongoing management and maintenance of the building. The parties involved in design delivery and operation of a building will need to consider how the integrity of the model can be assured throughout a building's lifecycle.
To address the issues outlined in this paper, there is a need for those involved in the design, construction and operation of intelligent buildings to have a good awareness of cyber security issues. Attempting to address the issues retrospectively on handover of a new building is rarely satisfactory and may involve significant rework and cost, while failing to deliver a resilient solution. Cyber security is as critical to intelligent buildings as disciplines such as structural and fire engineering are to multi-storey buildings. It will need to become part of the routine risk assessment process for all building projects, just as issues such as ground conditions, earthquake risk and so on are assessed based on the building location.
Applying appropriate systems and software engineering techniques to the design of the cyber environment is an essential first step. A number of these techniques are covered in the ‘20 Critical Controls’ , which should be a reasonable starting point for the cyber security protection of any intelligent building. There is a significant body of knowledge relating to the development of trustworthy software  and lessons that can be learned from the design and implementation of safety-critical systems.
With increasing use of network and information technology to deliver smart or intelligent buildings there are a number of technical and management challenges to address. The use of these technologies potentially offers significant benefit to society, but should be balanced with the inherent cyber security risks. If these risks are not addressed throughout a building's lifecycle there could be serious health and safety issues as a consequence.
There is a considerable body of knowledge available regarding systems and software engineering practice and their application to safety-critical systems. Applying this knowledge and combining it with best practice cyber security guidance such as the ‘20 Critical Controls’ could significantly reduce the vulnerability of intelligent buildings. The IET developed briefing document, referred to in this article is intended to help raise awareness.
The ACE industries urgently need to raise cyber security awareness, particularly with their increasing dependence on network and IT solutions. This awareness should cover the full building lifecycle and will need to be accompanied by an appropriate skill level for those involved.
The author wishes to acknowledge the support received from the UK Centre for Protection of National Infrastructure (CPNI) during the development of a report on Resilience and Cyber Security of Technology in the Built Environment .
- Fisk D.: ‘Cyber security, building automation, and the intelligent building’, Intell. Build. Int., 2012, 4, (3), pp. 169–181 (doi: 10.1080/17508975.2012.695277).
- Clements-Croome D. J.: ‘What do we mean by intelligent buildings?’, Autom. Constr., 1997, 6, (5), pp. 395–400.
- DEGW & Teknibank: ‘The intelligent building in Europe: executive summary’ (British Council of Offices, London, 1992) (Occasional Paper 3).
- Ehrlich P.: ‘What is an intelligent building’, 2005. Available at: http://www.automatedbuildings.com/news/aug05/articles/ibtpe.htm, accessed 10 June 2013.
- Harrison A.: ‘Intelligent building to the distributed workplace’, in Worthington J. (Ed.): ‘Reinventing the workplace’ (Architectural Press, Oxford, 2006, 2nd edn.), pp. 100–103.
- Boyes H. A.: ‘Resilience and cyber security of technology in the built environment’ (The Institution of Engineering and Technology, London, 2013), pp. 8–10. Available at: http://www.theiet.org/resources/standards/cyber-buildings.cfm, accessed 18 June 2013.
- International Air Transport Association: ‘Fact sheet: Industry statistics’, Canada, 2012. Available at: http://www.iata.org/pressroom/facts_figures/fact_sheets/Documents/industry-facts.pdf, accessed 10 June 2013.
- Purnell J.: ‘Information technology systems at airports: a primer’ (Transportation Research Board, Washington, DC, 2012), pp. 70–87.
- International Telecommunications Union: ‘Series X: data networks, open system communications and security, telecommunications security: overview of cybersecurity’ (I. T. U., Geneva, Switzerland, 2008) (ITU-T X.1205).
- Greene S. S.: ‘Security policies and procedures’ (Pearson Education, 2006).
- Norton R.: ‘Unintended consequences’, in HendersonD.R. (Ed.): ‘The concise encyclopaedia of economics’ (Liberty Fund Inc, Indianapolis, 2008, 3rd edn.).
- Boyes H. A.: ‘Resilience and cyber security of technology in the built environment’ (The Institution of Engineering and Technology, London, 2013), p. 17.
- Boyes H. A.: ‘Resilience and cyber security of technology in the built environment’ (The Institution of Engineering and Technology, London, 2013), pp. 18–19.
- Connett D.: ‘Four jailed after Brink's-Mat bullion trial’. The Independent, 1992. 18 August 1992. Available at: http://www.independent.co.uk/news/uk/four-jailed-after-brinksmat-bullion-trial-1540982.html, accessed 18 June 2013.
- Stewart J. M. Tittel E. Chapple M.: ‘CISSP: certified information systems security professional study guide’ (Wiley, Indianapolis, 2011, 5th edn.), p. 240.
- BIM Taskgroup: ‘BIM protocol – overview’, 2013. Available at: http://www.bimtaskgroup.org/bim-protocol, accessed 26 June 2013.
- Weiss J.: ‘Protecting industrial control systems from electronic threats’ (Momentum Press, New York, 2010), p. 34.
- Rios B.: ‘Google's buildings hackable’, 2013. Available at: http://www.cylance.com/techblog/Googles-Buildings-Hackable.shtml, accessed 26 June 2013.
- Symantec: ‘W32.Stuxnet’, 2010. Available at: http://www.symantec.com/security_response/writeup.jsp?docid=2010-071400-3123-99, accessed 26 June 2013.
- Kaspersky Lab.: ‘Kaspersky lab experts provide in-depth analysis of flame's C&C infrastructure’, 2012. Available at: http://www.kaspersky.com/about/news/virus/2012/Kaspersky_Lab_Experts_Provide_In_Depth_Analysis_of_Flames_Infrastructure, accessed 26 June 2013.
- Bencsáth B. Pék G. Buttyán L. Félegyházi M.: ‘Duqu: analysis, detection, and lessons learned’. ACM European Workshop on System Security (EuroSec), 2012, vol. 2012.
- Hickey A. R.: ‘Schneider electric: cloud computing key to power, energy management’, 2010. Available at: http://www.crn.com/news/cloud/227500884/schneider-electric-cloud-computing-key-to-power-energy-management.htm, accessed 26 June 2013.
- Sans Institute: ‘The critical security controls’, 2013. Available at: http://www.sans.org/critical-security-controls, accessed 26 June 2013.
- Trustworthy Software Initiative: ‘About the TSI’, 2013. Available at: http://www.uk-tsi.org.uk/?page_id=105, accessed 26 June 2013.