Impact case study database
Search and filter
Filter by
- The University of Kent
- 11 - Computer Science and Informatics
- Submitting institution
- The University of Kent
- Unit of assessment
- 11 - Computer Science and Informatics
- Summary impact type
- Health
- Is this case study continued from a case study submitted in 2014?
- No
1. Summary of the impact
Srivas Chennu and Howard Bowman have developed a software tool for improving the diagnosis of conscious awareness in patients after severe brain injury, based on their research into using neural signal processing methods for quantifying human brain networks. Since 2016 the software tool has been deployed across two residential care homes in Cambridgeshire and Hertfordshire, a world-leading European assessment centre in Belgium, and a rehabilitation centre in Germany. Across these locations, it has been used to measure awareness at the bedside of 200+ patients. The EEG software has improved assessments by allowing clinicians to form a detailed profile of a patient’s cognitive function, and as a bedside tool it has advanced care practice by averting the need for transportation, while also providing a digital visualisation of the brain’s activity, which improves understanding amongst patients’ families and carers. As highlighted by one clinician, the tool is ‘significantly reducing the cost of care and specialist assessments by many thousands of pounds, while simultaneously improving timeliness and efficacy’ [c].
2. Underpinning research
Overview
There is currently little reliable data on the number of patients in prolonged disorders of consciousness in the UK. Estimates vary widely, with between 4,000-16,000 patients diagnosed to be in a vegetative state, and three times as many in minimally conscious states. Because diagnosing and treating such patients poses a serious clinical challenge to neurorehabilitation specialists and clinicians, levels of consciousness can be misdiagnosed in as many as 40% of vegetative state patients.
Chennu and Bowman’s research is transforming how such patients are assessed, managed and diagnosed. The team began researching computational methods for quantifying human brain networks in 2006 [R1]. In 2010, Chennu took up a postdoctoral research post elsewhere, while Bowman continued the work at Kent [R2]. Following Chennu’s return to Kent in 2016, the pair continued to develop their research together, with support from a Kent team of a Research Associate and a PhD student. The work was also supported by a grant from the Engineering and Physical Sciences Research Council awarded to Chennu [G1].
Their research demonstrated that following non-invasive bedside measurements of electrical brain activity, computational analysis can uncover consciousness in patients who otherwise might seem unaware [R3, R4, R5, R6]. This research led Chennu and Bowman to develop a software tool for analysing electrical brain activity measured at the patient’s bedside. The Kent team worked in collaboration with clinicians at a range of institutions who since 2017 have deployed the software tool and collected feedback data to further improve the tool.
Development of the underpinning research
Chennu and Bowman’s research involves adapting neural signal processing methods for quantifying human brain networks. Human brain network activity is the basis for the quantification of consciousness and is central to the improvement of diagnosis and rehabilitation plans. The foundations for this research were published in 2009 [R1] and since then the research has been refined and elaborated and the software tool developed.
Since 2016, Chennu and Bowman have built upon this work, combining neural signal processing, machine learning, and computational analysis of electrical brain activity to characterise the interaction between attention and consciousness. The work was developed for clinical application by showing that analysis of electrical brain activity could uncover hidden awareness in some patients with so-called disorders of consciousness following severe brain injury, even though these patients might seem unresponsive at the bedside [R3].
This computational methodology was validated in three complementary ways: first, using phase synchrony for the identification of consciousness states between coma and quasi-brain-death [R2]; then using large-scale datasets of brain activity acquired from clinical settings [R3]; and, most recently, using brain activity recorded during anaesthetic-induced unconsciousness [R6]. In particular, [R6] characterised the brain network’s underlying transitions from full consciousness to reduced awareness. Across this network, brain-scale patterns of activation drive the signatures of awareness used, for example, in output [R3] to assess consciousness.
Application of the research through EEG software
Based on the findings of this underpinning research, between 2016 and 2017 Chennu and Bowman made advances towards developing a software tool for analysing electrical brain activity measured at the patient’s bedside to index their level of consciousness automatically, without requiring any behavioural responses from the patient.
The software tool involves using high-density electroencephalography (EEG) to measure electrical brain activity at the patient’s bedside; the software analyses these recordings and produces an assessment of the patient’s brain function and a digital visualisation of the brain’s activity. The results of the assessments can then be used to develop a detailed brain-based profile of the patient's state of awareness and cognitive function and to formulate an individualised care and therapy plan for the patient.
In 2018, the EEG software tool was deployed in a clinical trial of therapeutic intervention to aid the recovery of patients after severe brain injury [R4]. During the trial, the tool was used to measure the efficacy of electrical stimulation therapy in individual patients by assessing them with the software before and after the stimulation.
More recently, research has demonstrated the viability of routine deployment of the EEG software tool in clinical rehabilitation centres for repeatable and regular assessments of patients [R5]. This later work demonstrated that the software can be valuable for augmenting standard behavioural assessments of consciousness in the clinic, by diagnosing current levels of awareness and predicting longer-term recovery.
3. References to the research
Chennu, S., Craston, P., Wyble, B. & Bowman, H. ( 2009). ‘Attention Increases the Temporal Precision of Conscious Perception: Verifying the Neural-ST2 Model’. PLoS Computational Biology, 5 , e1000576. https://kar.kent.ac.uk/54653/.
Li, L., Witon, A., Marcora, S., Bowman, H., & Mandic, D. P. ( 2014). ‘EEG-based brain connectivity analysis of states of unawareness’. In 36th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, pp. 1002-1005. https://kar.kent.ac.uk/48649/.
Chennu, S., Annen, J., Wannez, S., Thibaut, A., Chatelle, C., Cassol, H., Martens, G., Schnakers, C., Gosseries, O., Menon, D. & Laureys, S. ( 2017). ‘Brain networks predict metabolism, diagnosis and prognosis at the bedside in disorders of consciousness’. Brain, 140 , 2120-2132. http://kar.kent.ac.uk/61849/.
Thibaut, A., Chennu, S., Chatelle, C., Martens, G., Annen, J., Cassol, H. & Laureys, S. ( 2018). ‘Theta network centrality correlates with tDCS response in disorders of consciousness’. Brain Stimulation. https://kar.kent.ac.uk/69096/.
Bareham, C. A., Allanson, J., Roberts, N., Hutchinson, P. J. A., Pickard, J. D., Menon, D. K. & Chennu, S. ( 2019). ‘Longitudinal assessments highlight long-term behavioural recovery in disorders of consciousness’. Brain Communications, 1, fcz017. https://kar.kent.ac.uk/76439/.
Witon, A., Shirazibeheshti, A., Cooke, J., Aviles, A., Adapa, R., Menon, D. K., Chennu, S., Bekinschtein, T. A., David Lopez, J., Litvak, V., Li, L., Friston, K., and Bowman, H. ( 2020). ‘Sedation modulates fronto-temporal predictive coding circuits and the double surprise acceleration effect’. Cerebral Cortex, bhaa071 . https://kar.kent.ac.uk/80568/.
Grants
- Engineering and Physical Sciences Research Council (EPSRC) grant EP/P033199/1 ( 2017-2019). ‘Reducing Misdiagnosis of Awareness after Severe Brain Injury’. PI: Srivas Chennu. Value £101,058.
4. Details of the impact
The EEG software tool was first deployed in 2016, on 40 patients at two residential care homes in Cambridgeshire (Askham Village Community, Doddington) and Hertfordshire (Gardens and Jacobs long-term residential rehabilitation centres in Sawbridgeworth) [c, e, f, g]. Since then, its rollout has gathered significant momentum, involving a further 160+ patients at the Coma Science Group in Liege, Belgium, ( 2017-present) [b]; and 26 patients at the Therapiezentrum Burgau rehabilitation centre in Germany ( 2018-present) [d].
The results of the EEG software assessments allow clinicians to form a detailed brain-based profile of a patient's state of awareness and cognitive function. This allows more tailored care and therapy to best meet the needs of the individual patients. The tool has made a significant impact on clinicians, patients and their families in the UK and Europe, by:
advancing the clinical approach to brain activity assessments,
improving severe brain injury rehabilitation and care practices, and
increasing understanding and involvement amongst patients’ families.
The co-director of Europe’s leading centre for clinical assessment, management and treatment of patients diagnosed with prolonged disorder of consciousness (Coma Science Group) explains how ‘The adoption of the technology has informed our knowledge, patients’ families understanding, and, most importantly, the overall care of the patients we treat’ [b].
Advancing clinical assessments
Feedback from clinical staff indicates that the EEG software tool has had a significant practical impact on the assessment of brain activity in severely ill patients. For example, the Professor of Anaesthesia at Addenbrookes Hospital has described how the software has ‘advanced the clinical practice of brain activity assessments, improving practitioners’ and specialists’ understanding of the recovery of consciousness after severe brain injury’ [c].
One of the UK’s leading clinical experts in neurorehabilitation has stated that not only is the technology ‘helping to inform clinical practice in real-life settings across several specialist hospitals in the UK and Europe, with the potential for more on the horizon’, but also noted that she has ‘benefitted from its value in allowing me and those I work with to more accurately quantify the level of activity associated with consciousness’ [a]. She highlighted how the research ‘has fostered new knowledge around how brain activity assessments can reduce the rate of misdiagnosis of consciousness’ and is informing ‘our current thinking for how to make the technology more readily available to patients and advance what we do in the future’ [a].
Advancing rehabilitation and care practices
Prior to the uptake of the EEG software tool, many of the patients in the Cambridgeshire care setting were treated at Addenbrooke’s during the acute period after an injury, before being transferred to the centre for prolonged rehabilitation. The transferral process made patient follow-ups more difficult as it required bringing them back for frequent assessments and intervention. In his supporting letter, the Professor of Anaesthesia from Addenbrooke’s explains how the software has ‘bridg[ed] this clinical and translational gap between acute and chronic stages of recovery after brain injury by facilitating a way to manage patients at their bedside and avert the need for transportation’ [c]. In addition, he states that: ‘Using Dr Chennu’s technology, clinicians have been able to identify the most useful time points during a patient’s rehabilitation when they can benefit from such scans, thereby significantly reducing the cost of care and specialist assessments by many thousands of pounds, while simultaneously improving timeliness and efficacy’ [c].
Describing the changes made to their clinical practice as a result of the EEG software tool, the co-director of the Coma Science group explains that ‘the technology’s capacity to assess patients alongside other brain-based assessments like magnetic resonance imaging [which] has improved and often guided the care of our patients, thereby also informing our clinical decision-making and its outcomes’ [b]. He states that ‘Dr. Chennu’s research and expertise […] is useful to us as clinicians in informing how we can respond to the challenges faced in caring for patients with brain injury, as the tool that his work has led to is both easy-to-use and also improves the processes of our care’ [b].
A survey of families of patients at two of the participating rehabilitation centres corroborates this impact on the care received by patients. Responses highlighted how ‘It is much better to have the EEG at the bedside it does not interfere with the care of my relative,’ and ‘No disruption is caused to [the patient’s] routine’ [h].
Similarly, a survey of clinical staff across the two centres established that the software tool allowed clinicians ‘to consider the patients conscious state in more depth and provide more beneficial interaction’ and ‘reduces risks for patients – risk of infection for transport/moving, pressure sores, more convenient, less disruption to patient’s normal routine’ [i].
Increasing families’ understanding and involvement
By including a digital visualisation of the brain’s activity in a way that previous methods could not, South Newton hospital’s clinical expert believes that the EEG software tool ‘improves both the clinician and patients’ family and carers understanding of a patient’s brain activity’ [a]. She confirms that the technology has ‘enabled me and my colleagues to more easily communicate the meaning and impact of these assessments to the patients’ family and carers. Further, it often increases the extent of their understanding of the patient’s diagnosis/situation’ [a].
This beneficial impact on families’ involvement and understanding is supported by clinicians at the Coma Science Group . Because the software has identified a minimal state of conscious despite a patient’s behaviour indicating an unresponsive wakeful state, for instance, not only have clinicians in Liege ‘then proposed to intensify care’ but also ‘in such contexts, together with the family [they] have used the findings to inform and discuss’ patient treatment options’ [b]. The technology, they argue, ‘enhances the capacity for collaborative thinking with patient families and is a valued port of information amongst all parties, as it is often combined with the knowledge that patients’ families have of the patient to establish a best cumulative purview of the whole situation’ [b].
5. Sources to corroborate the impact
Letter from the Consultant Neuropsychologist and Director of Clinical Development, Renovo South Newton Limited, UK, describing the impact that the Kent research has had on the role of bedside brain activity assessments in patients with disorders of consciousness (30 July 2020).
Letter from the Clinical Research Scientist and Co-director of the Coma Science Group, University of Liège, Belgium, describing the positive impact that the Kent research has had on improving the detailed assessments on patients referred for specialist evaluation by clinical centres across Europe (16 August 2020).
Letter from Professor of Anaesthesia, University of Cambridge, UK, describing the novel advance of technology from the Kent work for bridging clinical and translational gap between acute and chronic stages of recovery after brain injury (17 September 2020).
Letter from the Chief Physician and Head of Department, Therapiezentrum Burgau, Germany, describing the successful use of the Kent high-density EEG brain activity assessments on 11 patients (25 June 2019).
Letter from a General Practitioner, Sawbridgeworth Medical Services, UK, describing the positive response of Kent’s EEG brain activity assessments on 28 patients, as indicated in the survey data from the patients and their family, as well as from staff at Gardens and Jacobs Neurological Centres (22 March 2019).
Letter from the Operations Director, Askham Village Community, UK, providing support for conducting a patient study at Askham Village Community, UK (27 March 2015).
Letter from the Corporate Neuro Clinical Lead, Ramsay Neurological Services, UK, providing a supporting statement for conducting a patient study at Gardens and Jacobs Neurological Centres, Sawbridgeworth, UK (27 March 2015).
Report of a Family Questionnaire. To determine the acceptability and feasibility of the EEG healthcare technology developed, questionnaires were administered to consultees and families of patients assessed using the technology. Across two participating rehabilitation centres where patients were resident (in Sawbridgeworth and Doddington), 22 surveys were sent out and 11 patients’ family members completed them (date of survey letter: 23 February 2017).
Report of a Clinical Care Staff Questionnaire. To assess the clinical utility and feasibility of the EEG healthcare technology developed, questionnaires were administered to the clinical staff involved in the care and management of patients assessed using the technology. Across two participating rehabilitation centres where patients were resident (in Sawbridgeworth and Doddington), 17 of the 42 administered questionnaires were completed and returned (date of survey letter: 23 February 2017).
- Submitting institution
- The University of Kent
- Unit of assessment
- 11 - Computer Science and Informatics
- Summary impact type
- Technological
- Is this case study continued from a case study submitted in 2014?
- Yes
1. Summary of the impact
From 2004 to 2017, University of Kent researchers developed novel federated identity management (FIM) and privilege access management (PAM) technologies, which have been integrated in OpenStack since November 2013. OpenStack is a community cloud project involving 135+ international leading vendors, including Lenovo, Oracle, RedHat, IBM, Rackspace, and Cisco. FIM allows OpenStack users to use their existing usernames and passwords, instead of having to create a new pair whenever they want to access a new cloud service. Introducing FIM has been recognised by OpenStack for accelerating and shaping the project’s federated identity system’s performance and by OpenStack users for advancing their organisational security and efficiency whilst also saving significant costs.
2. Underpinning research
Background
Since 2004, Professor Chadwick and his team in the School of Computing at Kent have undertaken extensive research in the field of Information Systems security. The work began with the development of PERMIS – a suite of open-source privilege and access management (PAM) software that provides novel authorisation technologies to developers’ software systems. This was the subject of an impact case study in REF2014. Since 2009, the research has progressed into the novel application area of federated identity management (FIM).
The underlying problem addressed by the research was the hundreds of usernames and passwords that users are forced to remember, which leads to significant vulnerabilities. FIM greatly reduces this number by allowing a single identity provider (using one username/password) to log a user into dozens of federated systems. The FIM research has been of practical benefit to OpenStack – the most popular open source cloud software in the world that has a membership of 135+ international organisations, including global vendors such as Lenovo, Oracle, RedHat, IBM, Rackspace, Cisco, Dell, Fujitsu and Huawei. Prior to November 2013, OpenStack only supported traditional access control methods by individually allocating new usernames and passwords to users – a process that reduced security, impeded business proficiency, and wasted time for both users and service providers. The work on FIM changed that by allowing users to authenticate to an OpenStack cloud using their existing credentials through an external Identity Provider, rather than creating another username and password pair for a particular OpenStack cloud.
The research
Chadwick’s reputation and expertise in the field of system security led to him being invited to present the concept of FIM at FOSAD 2008. The International School on Foundations of Security Analysis and Design (FOSAD) is one of the foremost events for disseminating knowledge in the area of computer systems and networks security. In 2009, the presentation evolved into a seminal and highly-cited paper that provides an introduction to digital identity, identity management, and FIM, and the key issues faced, as well as ideas on how to address them [R1].
In 2011, following the work on PERMIS [G1] and the FOSAD talk, Chadwick responded to an ESPRC call for cloud computing with an argument for the integration of FIM. The resulting funded project [G4] led to the publication of an introductory paper [R2], which describes a policy-based authorisation infrastructure that can be used by cloud providers to ensure that the usersʼ privacy policies are stuck to their data. This was followed by a paper demonstrating the feasibility of a policy-based authorisation infrastructure in cloud infrastructure [R3], and a final results paper that describes open source software to support a trust, privacy and security infrastructure for the cloud [R4].
This research initiated Chadwick’s collaboration with OpenStack. He presented the approach at several OpenStack meetings between 2011 and 2013. Since there were many different FIM protocols in use and still in development, one of the key challenges was how to support interoperability. The concept, protocol independent FIM, was adopted by the OpenStack developers and now all OpenStack releases support FIM with several protocols supported as standard.
Furthermore, new FIM protocols can easily be added, as and when they are developed. The incorporation of FIM into OpenStack was demonstrated in 2014 in [R5], which presents a detailed federated identity protocol sequence. The paper also describes the implementation of the protocol independent system components, along with the incorporation of two different FIM protocols, namely SAML and Keystone. The FIM protocols were further validated through GÉANT's CLASSe project ( 2013-2015) [G6], which showed how cloud services such as OpenStack can be smoothly integrated in a federation built with an AAA (Authentication, Authorisation, and Accounting) infrastructure, by using Application Bridging for Federated Access Beyond Web (ABFAB) technologies [R6].
3. References to the research
Chadwick, D. ( 2009). ‘Federated Identity Management’. In: Aldini, A., Barthe, G. and Gorrieri, R. eds. FOSAD 2008/2009. Berlin: Springer-Verlag, pp. 182-196. http://dx.doi.org/10.1007/978-3-642-03829-7_3. https://kar.kent.ac.uk/30609/
Chadwick, D., Lievens, S., den Hartog, J., Pashalidis, A. and Alhadeff, J. ( 2011). ‘My Private Cloud Overview: A Trust, Privacy and Security Infrastructure for the Cloud’. In: 2011 IEEE 4th International Conference on Cloud Computing, pp. 752-753. http://dx.doi.org/10.1109/CLOUD.2011.113. https://kar.kent.ac.uk/43198/
Chadwick, D. and Fatema, K. ( 2012). ‘A privacy preserving authorisation system for the cloud’. Journal of Computer and System Sciences, 78(5):1359-1373. ISSN 0022-0000. https://doi.org/10.1016/j.jcss.2011.12.019. https://kar.kent.ac.uk/31975/
Chadwick, D., Casenove, M. and Siu, K. ( 2013). ‘My private cloud--granting federated access to cloud resources’. Journal of Cloud Computing, 2:1-16. http://dx.doi.org/10.1186/2192-113X-2-3. https://kar.kent.ac.uk/43204/
Chadwick, D., Siu, K., Lee, C., Fouillat, Y. and Germonville, D. ( 2014). ‘Adding Federated Identity Management to OpenStack’. Journal of Grid Computing, 12:3-27. http://dx.doi.org/10.1007/s10723-013-9283-2. https://kar.kent.ac.uk/43212/
Pérez Méndez, A., López Millán, G., Marín López, R., Chadwick, D. and Schechtman Sette, I. ( 2017). ‘Integrating an AAA-based federation mechanism for OpenStack–The CLASSe view’. Concurrency and Computation: Practice and Experience, 29 (12). Article Number 4148. ISSN 1532-0626. http://dx.doi.org/10.1002/cpe.4148. https://kar.kent.ac.uk/61206/
Awards and Grants
SIPS grant ( 2004-2006). ‘Seamlessly Integrating PERMIS and Shibboleth’. PI: David Chadwick. Value £61,000.
DyCom grant ( 2004-2007). ‘Multilayer Privilege Management for Dynamic Collaborative Scientific Communities over Environmental eScience Data Grids’. PI: David Chadwick (Joint bid with CLRCC). Total Value: £185K (£77,000 for Kent).
DyVOSE grant ( 2004-2006). ‘Dynamic Virtual Organisations in e-Science Education’. PI: David Chadwick (Joint bid with National eScience Centre Edinburgh, University of Glasgow, EDINA). Total Value: £283K (£107,500 for Kent).
EPSRC EP/I034181/1 ( 2011). ‘My Private Cloud’. PI: Chadwick. Value £57,808.
EPSRC EP/J020354/1 ( 2012-2013). ‘Sticky Policy Based Open Source Security APIs for the Cloud’. PI: Chadwick. Value: £126,939.
EC GÉANT grant (2013-2015). ‘Cloud-ABFAB Federation Services in Eduroam (CLASSe)’. PI: David Chadwick. Value: €89,000.
4. Details of the impact
Advancing OpenStack’s commercial innovativeness and efficiency
Chadwick was invited to the OpenStack summits in the USA in 2012 and 2013 to discuss the Kent research on FIM, educating and informing OpenStack stakeholders on its processes through the discussion of key blueprints [a]. As a result, in November 2013 Chadwick’s novel FIM design was implemented in the ‘H’ phase of OpenStack, adding federated access to Keystone (OpenStack’s identity manager) [a][b]. The Kent FIM protocol was the first (and remains the only) system to have introduced federated identity to OpenStack, and since its implementation in November 2013 Chadwick has continued to support OpenStack and its developers and users via his advisory role in the Keystone working group ( 2011-2017).
In response to Chadwick’s continued contribution and support of the Kent FIM, OpenStack’s Keystone Project Lead, Steve Martinelli, stated: ‘David [Chadwick] and his team[‘s …] expertise and guidance in the area of Federated Identity has been priceless. I can confidently say that their collaboration and background on the subject has helped accelerate and shape Keystone's support of Federated identity […] David's opinion is sought out when discussing Keystone issues on OpenStack's mailing list, where his deep understanding of security, and identity management is made evident. The same can be said about David's involvement at OpenStack summits (our bi-annual design conferences), where he is involved in the architectural design of new Keystone features’ [c].
Jisc – a company supporting institutions of higher education and research in the UK – also benefits, as stated by Stefan Paetow, Jisc’s Research Liaison Coordinator: ‘University of Kent readied the Openstack software stack for use with federated identity technologies, including Moonshot, by piloting Openstack with the pre-Assent Moonshot pilot infrastructure and the UK Access Management Federation, submitting fixes to Openstack and Jisc where required, and driving the web-based use case for Moonshot forward’ [d].
Improving OpenStack developer proficiency
Many developers are interacting with OpenStack in order to use its interface. Of these developers, Chadwick has advised and educated many as a result of his consultancy via the OpenStack summits and working group mailing list. His Kent team has also generated many new lines of code or patches, for instance, to enhance the usability and proficiency of the software. A senior manager at RackSpace identified that between 2011 and 2014: ‘University of Kent has contributed 1400 lines of code, 102 reviews, 71 patches, 14 blueprints [and…] have been approachable subject matter experts, whose knowledge was widely adopted by the Keystone community’ [e]. Similarly, an engineer from RedHat in 2017 stated that: ‘Professor Chadwick was essential in enabling the project to properly implement Identity Federation, a set of features without which OpenStack would be unusable by a great many organizations […]. As a software engineer at Red Hat, I have seen a direct correlation between the success of our implementation and the work that Professor Chadwick and his team contributed to the project’ [f].
Enhancing OpenStack users’ security and cost-effectiveness while reducing administrative burden
Users save time and effort while increasing security, because with FIM, users do not have to create and remember a large number of different passwords. Having to create and memorise numerous passwords leads users to choose weak and easily crackable passwords. With FIM, they need to create and maintain only one strong password.
Over 135 international organisations are currently utilising OpenStack, and running their jobs and applications off it. We carried out a survey of OpenStack users to assess the impact of having FIM integrated into Open Stack [g]. Sixty per cent of the respondents (n=14) stated that adopting FIM makes the system more (or much more) secure. A Cloud Manager from a national computer network for universities and research noted: ‘[we] don’t have to manage safety of credentials, since we store none. We are not subject to brute-force, since services are accessed via identity tokens only’ [g]. The Head of Identity Management of the same institution noted further: ‘No passwords inserted in web forms, no passwords kept in the keystone, no recovery/force-passwords-on-the-phone, etc.’ [g].
Traceability is important as part of implementing security, and this is facilitated by FIM. A System Administrator at a small company highlights the benefit of FIM: ‘We will get to know who accesses what and when’, while a researcher pointed out: ‘R&S and Sirtfi [Security Incident Response Trust Framework for Federated Identity] helps. Also using the institutional accounts helps with traceability’ [g]. Similarly, an organisation’s security relies on a proper audit of their users. For example, it is imperative to revoke access from former employees who had left the organisation. Evidently, the advantage of FIM is clear, as further established by one Network Engineer: ‘We can audit access controls from a central system and we can easily require MFA (Multi Factor Authentication) for user logins’ [g]. This is echoed by an Administrator/Architect from an American telecommunications company dealing with several thousand users: ‘Old account remove quicker’; this echoes the view of a Technical Architect at another organisation, who states: ‘revocation is central. Centralized group management assignment are key as well’ [g].
In addition, FIM offers a potential for saving time and money. The time-saving benefit of having FIM is highlighted by a Cloud Manager from a European institution: ‘Registration time is just few clicks wrt [with respect to] alternative procedure for identity verification’ [g]. An Administrator at a large American telecommunications company elaborated further: ‘Multiple Operation people were spending >30% of their time just managing accounts before we implemented hooking into corporate Active Directory System’ [g]. Meanwhile, a Cloud Architect at a UK Government agency believed that FIM reduces operational costs, affording a saving of ‘0.5 of a person to maintain the identity proxy’ [g]. Similarly, a System Administrator at a small company noted a saving of ‘roughly $5000’, while the Head of Identity Management at a European institution reported a saving of ‘0.2 FTE’ [g].
Administrative burden can also be reduced by using FIM, with 47% (n=11) of the respondents suggesting that there is less or much less administrative burden in running their system. A Senior Software Engineer at a public open cloud exchange project commented: ‘A unified authentication management has allowed us to let users use the same accounts for OpenStack and OpenShift, which we are operating. So we and users don't have to deal with multiple accounts or systems of access management […] Also we can offload verifying a user to the identity providers they are coming from’ [g]. A Technical Specialist at a European organisation pointed out: ‘when a user leaves the home organisation, and his account is deactivated, this user cannot log in anymore in the open stack environment. Effectively, his permissions are revoked while openstack maintainers don't know (and now don't care) that he is gone’ [g]. A Cloud Platform Architect at a multinational telecommunications company summarised the impact in reference to his 3,600 users: ‘No more account management, users shall do it for us’ [g].
Based on his FIM research and as a result of the growing recognition of his expertise in this field, Chadwick was an invited expert to the nascent W3C Verifiable Credentials working group in 2017. His contribution led to his co-authorship of the Recommendation and its associated Implementation Guidelines [h] and Data Model [i]. In 2019, the University created a spin-out company, Verifiable Credentials Ltd, to capitalise on the implemented Proof of Concept.
5. Sources to corroborate the impact
Keystone/Federation/Blueprint Web-clipping: Presents the final version of the blueprint developed following the OpenStack Summit and describes how to add protocol independent federated access to Keystone, the identity service of OpenStack. https://wiki.openstack.org/wiki/Keystone/Federation/Blueprint
Presentation by CERN at the OpenStack summit in Hong Kong in Nov 2013. Slide 24 reveals the pathway to impact that Kent has had on CERN's efforts by collaborating with them. https://www.slideshare.net/andrewrhickey/cern-rackspaceopen-stacksummit
Letter from OpenStack’s Keystone Project Lead at IBM, describing Chadwick’s continued contribution to help accelerate and shape Keystone's support of Federated Identity (26 November 2015).
Letter from Moonshot Industry & Research Liaison Coordinator at Jisc, describing the University of Kent’s contribution in vital resources and research to the development and testing of the Moonshot technology (14 January 2016).
Letter from the Senior Manager, Product at Rackspace, describing the contributions of Chadwick’s team to improving OpenStack developers’ proficiency (30 November 2015).
Letter from Cloud Solutions Architect at Red Hat, Inc., describing the essential role played by Chadwick in enabling OpenStack to properly implement Federated Identity (15 November 2017).
Survey data of OpenStack Stakeholders, based on input from 23 stakeholders, including service providers and service users (23 December 2019 to 31 January 2020).
W3C Verifiable Credentials Implementation Guidelines 1.0: Implementation guidance for Verifiable Credentials (24 September 2019). https://www.w3.org/TR/vc-imp-guide/
W3C Verifiable Credentials Data Model 1.0: Expressing verifiable information on the Web (19 November 2019). https://www.w3.org/TR/vc-data-model/
- Submitting institution
- The University of Kent
- Unit of assessment
- 11 - Computer Science and Informatics
- Summary impact type
- Societal
- Is this case study continued from a case study submitted in 2014?
- Yes
1. Summary of the impact
Pedagogical and technical research in computer science at Kent led to the development of Greenfoot – a unique software system that supports the learning of programming at school level (age 13+). Greenfoot has generated reach in 49 countries, been translated into 12 languages, and, since August 2013, has improved the learning of programming for more than 3.9 million schoolchildren and supported the teaching capacity and innovativeness of over 5,000 teachers. Additionally, Greenfoot has been used by Computing at School (CAS) and the Welsh Joint Education Committee (WJEC) to develop and improve the provision of computer science through the national curriculum, and employed by key stakeholders, such as Oracle and Technocamps, to advance the profiling of programming and drive knowledge. For instance, the Oracle Academy Greenfoot curriculum was studied by thousands of students in over 70 countries in 2020.
2. Underpinning research
Background
Traditional approaches to programming education used to rely on text-based development environments designed mainly for professional programmers. From a student-pedagogical perspective, these environments are difficult to use, hard to interact with, and force learners through large amounts of detail before enabling them to achieve the first motivating impact. Using such programmes in secondary/high-school education settings, therefore, leads to problems in comprehension, engagement, and motivation among the students; they become resistant to learning programming despite software usage and understanding of its fundamentals being an ever more important, ingrained skill set applied in society. Similarly, teachers also found the text-based programming environments problematic for delivering modern programming paradigms in an innovative and engaging way. Responding to these issues, from 2005 to 2017 a Kent project team led by led by Kölling, and comprising Utting, Brown, Henriksen, and McCall, developed Greenfoot – an integrated development environment that allows easy conception of two-dimensional graphic applications – along with other supporting, instructional user websites and easy-to-use material.
Research and its pathways to impact
The Greenfoot project developed from two strands – pedagogical research and technical computer science research. The project provided an introduction to the different elements of the ecosystem (community, discussion, availability of material), along with the conception of the Greenfoot textbook [R6]. The Kent project group carried out research in both areas, including research in the development of new interaction techniques to support the learning of programming; the design and implementation of the software; and, the usability and effectiveness of the resulting system.
In 2005, the initial phase of pedagogical research and system design was led by Kolling and carried out jointly with Henriksen and John Rosenberg and Bruce Quig from La Trobe University. Then, in 2007, members of the Kent group (Brown, Kölling, McCall, Utting) made original contributions to Human Computer Interaction (HCI) in the context of novice programming systems, introducing interaction techniques that allow more experimentation and exploration, accelerate feedback, and, through this, increase engagement and motivation. This later work enabled the inception of new learning approaches that invert traditional curricula, such as allowing discussion of high-level ‘advanced’ concepts before discussion of syntax. With regard to programming environment technology, the research identified novel interaction techniques that enable a more direct and experimental approach to interacting with a programming system. Collectively, both aspects of this pedagogical strand of research were based on constructivist learning theories, which were specialised to the initial learning of programming, and, specifically, to the learning and teaching of concepts of object orientation. As a result, the understanding of key requirements to support initial learning of programming led to the formulation of goals and targets for system characteristics, such as interactivity, visualisation, and support for self-directed experimentation, thus making original contributions to computing pedagogy [R1, R2].
The second strand of the research was concerned with developing computer system abstractions and interaction techniques to realise the system characteristic goals in the context of a modern, statically typed, object-oriented language (Java, in this instance). Some of the techniques integrated into Greenfoot were modified from earlier highly interactive programming systems, such as Smalltalk and Self, and were adapted to the statically typed nature of Java. Specifically, the Kent group developed a new conceptual framework of object-oriented programming concepts and concrete novel interaction techniques [R3]. Then, in 2010, further research identified that the ecosystem has a major influence on learning success and the team sought to expand this by creating a community and social support for learning. As such, the Kent group created a technology-supported online community around Greenfoot, including the development of novel social interaction models, combining aspects of resource repositories [R2] and social networks [R4, R5].
3. References to the research
[R1] Kölling, Michael and Henriksen, Poul ( 2005). ‘Game Programming in Introductory Courses with Direct State Manipulation’. In: ITiCSE 2005: Proceedings of the 10th annual SIGCSE conference on Innovation and Technology in Computer Science Education. ACM Press, New York, New York (USA), pp. 59-63. ISBN 1-59593-024-8. https://dx.doi.org/10.1145/1067445.1067465. http://kar.kent.ac.uk/14304/
[R2] Fincher, Sally and Kölling, Michael and Utting, Ian and Brown, Neil C.C. and Stevens, Phil ( 2010). ‘Repositories of teaching material and communities of use: nifty assignments and the greenroom’. In: Proceedings of the Sixth international workshop on Computing education research. https://doi.org/10.1145/1839594.1839613. http://kar.kent.ac.uk/30638/
[R3] Kölling, Michael ( 2010). ‘The Greenfoot Programming Environment’. ACM Transactions on Computing Education (TOCE), 10(4):1-21. ISSN 1946-6226. https://doi.org/10.1145/1868358.1868361. http://kar.kent.ac.uk/30614/
[R4] Henriksen, Poul and Kölling, Michael and McCall, Davin ( 2010). ‘Motivating programmers via an online community’. J. of Computing Sciences in Colleges, 25(3): 182-196. ISSN 1937-4771. https://dl.acm.org/doi/abs/10.5555/1629116.1629132. http://kar.kent.ac.uk/30701/
[R5] Brown, Neil C.C. and Kölling, Michael ( 2013). ‘A Tale of Three Sites: Resource and Knowledge Sharing Amongst Computer Science Educators’. In: Ninth Annual International Computing Education Research Conference (ICER), Aug 2013, San Diego, USA, pp. 27-34. https://doi.org/10.1145/2493394.2493398. http://kar.kent.ac.uk/37642/
[R6] Kölling, Michael ( 2016). Introduction to Programming with Greenfoot: Object-Oriented Programming in Java with Games and Simulations, Second edition, Pearson. ISBN-10: 013-405429-6, ISBN-13: 978-013-405429-2. https://www.greenfoot.org/book/
Grants
The research was supported by a series of American industry grants over a period of several years, including: $650,000 from SUN Microsystems (2005-2008), $450,000 SUN foundation (2009-2010), $2m from Oracle Corp (2011-2017) and $36,000 from Google (2011-2013), totaling more than US$3.1 million.
4. Details of the impact
Greenfoot is supported by its own channel on YouTube with instructional videos (4,850 subscribers and over 1 million views) [a] and two supporting websites, one for its end users, the students (40,000 subscribers and 200,000 page views per month) [b] and one for teachers entitled the ‘Greenroom’ (5,419 subscribers, 241 resources and 4,114 discussion posts) [c]. Greenfoot has been adopted in over 49 countries [d], the user interface translated into 13 languages (including Chinese, Portuguese, German, and Spanish), and the Greenfoot tutorials translated into seven languages (Brazilian, German, Greek, Indonesian, Italian, Russian and Spanish) – with all translations produced by volunteers [b].
Fig. 1: Total unique users of Greenfoot (from [d])
Subsequently, the international proliferation of Greenfoot across the computer science education community has generated diverse forms of pedagogical impact. Since August 2013, the software has improved the learning of programming and benefitted more than 3.9 million school students directly [d, e, Fig. 1]; enhanced the teaching capacity of over 5,000 teachers who have used it to improve how they teach programming innovatively [c, e]; influenced national curriculum design in the UK; and, informed the practices, philanthropic missions, and opportunities of key computer science stakeholder groups, including CAS, WJEC, Oracle and Technocamps [f, g, h, i].
Improving student learning and engagement internationally
Unlike the traditional programming platforms used in schools, the usability and accessibility of Greenfoot enables students to achieve more tangible results more quickly, and, through the increased motivation and satisfaction that this fosters, enhances students’ understanding of programming concepts. In commenting on how Greenfoot increased learners’ motivation and engagement, a teacher from Sutton Grammar School (UK) highlighted that ‘My pupil was never a good student until I taught him Greenfoot. He got so inspired that he asked his parents to buy him the book, read the whole thing, and shot up to the very top of the year group in my subject with exam scores averaging 90%. He will be in my GCSE course next year, and I could not be more proud!’ [e]. Similarly, a teacher from the Jesuit High School in Oregon (USA) noted: ‘This amazing program motivates my students to learn object-oriented programming, including inheritance, interfaces, and the use of list and array structures, in an engaging and very fun manner’ [e].
Additionally, as pointed out by other teachers from across the USA and Germany, Greenfoot also boosts creativity and discovery in student learning: ‘My students love developing games and discovering the true creativity that is required to design and develop fun and interesting software’ and ‘My pupils can be as creative as they want’ [e]. Such increased motivation and creativity nurtures greater enthusiasm in the subject of Computer Science, as another UK teacher contends: ‘I get more enthusiasm teaching Greenfoot than any other aspect of teaching from year 7 to year 13. I believe that the reason my GCSE programme is so popular is due to what we teach in year 9. Out of 135 students in a year group, 63 have opted for Computer Science next year’ [e].
Enhancing teaching capacity and pedagogy innovativeness
Greenfoot [b], its supporting textbook (16,000 copies sold) [R6], and the Greenroom (the aforementioned teachers’ community and support site) [c] also improve programming teaching practice and capacity. In contrast to the disengaging environments and language of traditional methods, Greenfoot enables teachers to teach software development principles and object orientation in a modern, graphical context. As a teacher from Vincent-Lübeck-Gymnasium (Germany) observes: ‘The pedagogy behind the development environment has continuously been improved over time and on occasions made massive jumps forward. I can produce all my teaching tools inside Greenfoot [and] hence encourage the pupil to dissect my project structures and code’ [e]. Further to increasing innovation in programming pedagogy, teachers note that their job satisfaction and ability to teach inspirationally is improved through using Greenfoot, as a teacher from Carl-Helbing-Schule Emmendingen (Germany) testifies: ‘I often experience situations when the lesson is over, but some students still stay on longer, creating their games. Is there anything more satisfying an educator can dream of? The concepts of object-oriented programming become so much more clear and understandable for the students. It makes teaching concepts of OOP absolutely effortless’ [e]. Comparably, beyond the Greenfoot site itself and in the context of the Greenroom, teachers remark on how the supporting material and community sites enhance their ability to use Greenfoot and connect with other teachers. This point is made by a teacher from Ann Arbor Huron High School in Michigan (USA): ‘The nicest feature of it, beyond ease of use, is that there is a textbook that is built around the program! I have benefited from BOTH the blueroom and greenroom discussion boards. Getting lessons, questions answered, etc. The support for the product has been nice!’ [e]. This view is echoed by the teacher from Vincent-Lübeck-Gymnasium (Germany): ‘Being able to exchange knowledge with a global community of people who use Greenfoot in radically different environments is a huge help’ [e].
Informing UK computer science curricula
Since 2008, Computing at School (CAS) has collaborated with the Greenfoot team to accomplish its mission of promoting the teaching of computing at school across all school phases and supporting computing teachers with training and resources. Greenfoot is applied within several courses they run for teachers and since 2013 ‘has been an integral part of the Welsh Joint Education Committee (WJEC) GCSE curriculum’ [f]. The latter was part of the qualifications and reform programmes in England and Wales which saw the WJEC – a Welsh exam board – develop a new GCSE computer science qualification of which Greenfoot forms a part. In commenting on the benefits of Greenfoot’s integration into the GCSE, the WJEC highlight that the software ‘is used to develop an understanding of object orientated Programming using Java such that students can understand […] the concepts of inheritance and encapsulation’ [g]. The WJEC ‘value the use of Greenfoot as a component of the WJEC GCSE computer science qualifications, and the Greenfoot materials provide valuable resources for teaching’ [g].
In the same vein, the Vice-Chair of CAS also highlights that ‘across the last decade of computing resurging in UK schools, Greenfoot has contributed to the teaching of programming in school’ and ‘evolved to have a special place with pupils at KS4 and A level’ [f]. CAS testifies that ‘access to the innovative software encourages pupils to program: not only enhancing their interest in programming, but also enriching their understanding of computer science. The software provides an easily-understood interactive user interface, thus lowering the entry barrier to writing and understanding code, and has an active, supportive, worldwide community, tutorials, and teaching plans’ [f]. Correspondingly, for teachers ‘it has advanced the knowledge, skills and understanding of those teaching in the classroom’ [f]. CAS remarks in particular upon the innovativeness of Greenfoot in improving learning and teaching practice, asserting that ‘the pedagogical appeal and significance of Greenfoot in advancing the knowledge of and accessibility to computing is felt by both teachers and pupils; its technology has advanced the learning of staff and students in diverse ways’ [f].
Contributing to computer science philanthropy and NGO engagement initiatives
Oracle is an American multinational computer technology corporation headquartered in California that sells database software and technology, cloud-engineered systems, and enterprise software products. Since 2013 – as part of Oracle’s philanthropic mission to provide a wide range of free resources and works with institutions, educations and partners across the globe via their Oracle Academy education suite – Greenfoot ‘has been an important part of Oracle’s Academy’s curriculum’ [h]. As the Vice-President of Oracle confirms: ‘In the 2020 academic year to date, Oracle Academy Greenfoot related curriculum has been studied by thousands of students in over 70 countries’ [h]. Greenfoot, they argue, along with BlueJ and Stride, ‘collectively have the potential to bring programming to every student – not just those who aim for a career in software development or computer science’ [h].
Technocamps is a non-profit organisation started in 2003 with a mission to inspire, motivate, and engage people with computational thinking and promote Computer Science as underpinning all aspects of modern society. Technocamps have delivered more than 2,500 workshops to over 60,000 young people across Wales on programming, games development, and much more, and Greenfoot has been included as part of their workshop curricula. The Director of Technocamps states that: ‘in 2012 we developed our own workshop based on the Greenfoot material, and the reach of this workshop has increased substantially since 2013. Our Greenfoot workshop has been delivered to approximately 3,500 pupils over the past seven years across Wales, with 125 teachers undertaking CPD courses which include Greenfoot’ [i]. The workshop they go on and the interaction with the Greenfoot platform have ‘proven to be popular and valuable in influencing teaching pedagogy and student understanding […] Feedback from the teachers has suggested that the workshops are very beneficial for improving student learning and helped the teachers to guide their pupils through the GCSE’ [i]. Technocamps state that they have: ‘benefitted greatly from the Greenfoot software in supporting teachers and other classroom practitioners to deliver workshops on digital education and programming to young people in Wales’ [i].
5. Sources to corroborate the impact
Greenfoot YouTube channel, providing video tutorial and samples for using Greenfoot. http://www.youtube.com/user/18km
The official home page of Greenfoot: https://www.greenfoot.org/home/
The Greenroom community site for teachers using the Greenfoot software. To see map of subscribers in 50+ countries, log in with guest@example.com / guestpassword. https://greenroom.greenfoot.org/
Greenfoot user statistics (as of 16 December 2020). https://bluej.org/newstats/
Survey from Greenroom members (conducted between 21 June and 15 July 2019), with key statements and testimonials from teachers compiled into a report available online at: https://stewardship.atavist.com/bluej-greenfoot-report#chapter-5184004
Letter from Vice-Chair of Computing At School (CAS), describing Greenfoot’s continued contribution to the teaching of programming in school, and the benefits Greenfoot brings to teachers and pupils (14 October 2020).
Letter from WJEC Subject Officer Computer Science with ICT, describing how Greenfoot influences the specification of the WJEC GCSE computer science qualifications (6 November 2020).
Letter from the Vice-President of Oracle, describing Greenfoot’s contribution to preparing those who aim for a career in software development or computer science (4 December 2020).
Letter from the Director of Technocamps, describing how Greenfoot contributes to the digital and programming education of young people in Wales (21 October 2020).