Advanced Arterial Hypotension Adverse Event prediction through a Novel Bayesian Neural Network (AvertIT): was funded for 3-years (2m€) through the EU FW7 in 2008. The goals of AvertIT were to develop bed-side brain trauma systems that allow prediction of hypotensive events across 6 major neurological centres across Europe. Through integrating a variety of heterogeneous clinical data sets including real-time patient monitoring data sets (blood pressure, ECG etc), demographic data sets and laboratory/treatment data, and exploiting algorithms that have been trained through novel Bayesian algorithms, it is expected that hypotensive events in brain trauma patients can be identified with a 15 minute time window. This project undertook a large-scale multicentre clinical trial to judge the validity of the models put forward for adverse event prediction. A paper describing the output of AvertIT is given here.

Investigation of the molecular pathogenesis and pathophysiology of Disorders of Sex Development (EuroDSD): was funded for 3-years (3m€) through the EU FW7 in 2008. EuroDSD established a clinical data registry and research environment to support all aspects of rare sex diseases. The EuroDSD project included clinical partners from across Europe and has (up to now!) recruited over 1300 DSD cases. The technologies and standard operating procedures and processes for data collection, data access and usage in EuroDSD were widely endorsed by independent ethic oversight committees. Importantly many non-funded clinical and research collaborators are involved in EuroDSD. A paper describing one of the outputs of EuroDSD is available here.

Scots Words and Placenames (SWAP): aimed to engage the Scottish public in talking about the Scots words that they use and hear around them. The project sought any examples of words, their meanings, how they were used, where they were used, how they were pronounced, if people knew what they meant, whether they appeared on maps or were known through word-of-mouth; even how they looked (through uploading pictures). The results of the SWAP project were added to the word collections of Scottish Language Dictionaries and helped to form new dictionaries of the Scots language. They also contributed to knowledge of Scots place-names. The information we gathered on place-names was used to populate a comprehensive glossary of Scots place-name elements and to supplement the dictionary-based research, which was used to create it. This was an early example of a crowd-sourcing project.

Parliamentary Discourse (ParlDisc): focused on the extension of the ENROLLER systems to include access to the Hansard collections. The Hansard collections are the edited records of all parliamentary debates, votes, written ministerial statements and answers from the Houses of Commons and Lords. Its purpose is to provide day-to-day and accurate accounts of the proceedings in Parliament in order that they might be freely and easily accessed by any member of the public.

Real time River Pollution Identification using Animal Behavioural Models (ALARM) was funded ($560k) by the Victorian Department of Business and Innovation as part of the Broadband Enabled Innovation Program in 2011. Together with environmental researchers at the University of Melbourne, the PRESS team established the software and hardware infrastructure for real-time data acquisition of pollution events across the waterways of Victoria. This required a unique hardware/software solution that included video capture/processing, water quality sensor detection information (on water quality) and geospatial analytics on the associated streamed data to identify potential pollution events. The system was unique in that it had to work “in the field” without dedicated power and often without Internet. A paper describing this system is available here.

Meeting the Design Challenges of nanoCMOS Electronics (nanoCMOS): The nanoCMOS project WAS a £5.3M UK e-Science pilot project that began in 2006 that focused on how the electronics domain could gain benefit through the Grid. The intention of the project was to address the research challenges in dealing with transistors of increasingly small dimensions where atomistic effects cannot be ignored, and supporting a much wider transistor design space. Key aspects of this project included intellectual property protection. The electronics domain demands protection of intellectual property of data, designs and processes. Similarly, the software used in this domain was extremely expensive and typically license protected. An example paper from nanoCMOS is available here.

Research Data Storage Infrastructure (VicNode), funded ($3.2m) by the federal government Department of Education (it was formerly funded by the Department of Innovation, Industry, Science (also formerly known as DIISRTE/DIISR) through the Education Investment Fund (EIF) and the National Collaboration Research Infrastructure Strategy (NCRIS) activities. VicNode provides large-scale data storage systems that underpin a range of research activities across Australia with specific focus on Victorian researchers. The facility has been established and managed by the University of Melbourne and Monash.

National e-Infrastructure for e-Social Science (NeISS): was funded for 3-years in 2009 by the JISC. NeISS developed a research environment for social scientists interested in the area of social simulation. This can include for example future models of immigration or life expectancy on transport or infrastructure at local or national levels. Typical data sets include individual level census data, geo-spatial data and clinical data. Many of these data sets were under strict access and usage constraints, e.g. due to licensing arrangements or due to individual level patient information. The work of NeISS included support for large-scale data integration techniques and fine-grained security through user-friendly portals. An example paper from NeISS is available here.

Data Management through e-Social Science (DAMES): was funded in 2008 for 3-years by the ESRC. DAMES developed a specialist research environment for social science research. The social sciences are not unlike the clinical sciences in the volume and heterogeneity of distributed data sets that are being created. DAMES focused in particular on education, occupational, minority/ethnicity and on e-Health data sets. Particular studies included research into in the area of depression, self-harm and suicide. The data sets associated with this research covered the clinical sciences (hospital admission data; mental health data sets; primary care GP data sets); social science data sets including the UK Census and a variety of individual level microdata sets from providers such as the Office of National Statistics; and geospatial data sets from commercial providers such as the Ordnance Survey. The DAMES systems were validated and required ethics permission to support a wide range of clinical data linkage scenarios has been obtained. An example paper from DAMES is available here.

Modelling, Inference and Analysis for Biological Systems up to the Cellular Level was funded by the BBSRC in the UK and provided a high performance computing resource used to meet the computational demands of the bioinformatics/biological community.