New partnership for environmental data-sharing and analysis: 'champions' required

DAFNI is building a common platform for data storage and analysis to encourage a smarter, more ‘joined-up’ system for infrastructure planning

The DAFNI team (Data and Analytics for National Infrastructure), part of STFC Scientific Computing (SCD), is making strides in an effort to bring together researchers, organisations and authorities for a joined up approach to data analysis for national infrastructure. 

A DAFNI partnership that could bring environmental data-sharing to a new level is with CEDA (Centre for Environmental Data Analysis). CEDA and SCD jointly manage JASMIN, that stores vast quantities of global environmental data records. JASMIN is a ‘super-data-computer’ with storage and networking technologies dedicated to studying the natural environment. An exciting prospect is to link JASMIN and DAFNI to consider how infrastructure might be affected by changes in the environment.

The Science and Technology Research Council (STFC) are also inviting application for the role of Champions for DAFNI. DAFNI is currently under development and now has a “bronze” release that is suitable for test and evaluation by selected users to exercise the power of the platform and influence its ongoing development.

Champions will work with the DAFNI team and their home institutions to disseminate and promote the DAFNI platform within their institution, to explore use cases using the DAFNI facility within their research, and to feedback to DAFNI team on system requirements. Champions’ costs will be funded from the DAFNI project, and will involve access to the DAFNI platform and assistance from members of the DAFNI STFC team.

Brian Matthews is the DAFNI Group Leader. He says: 'Engineering researchers looking at infrastructure are finding they can’t make their models bigger or more complex to better reflect real life as they are limited by the computers in their labs. And better models need more data, which can be hard to share and access across organisations, especially when it needs to be handled in confidence. These were strong drivers behind the creation of DAFNI. Our team of dedicated developers has produced a system offering enough computing power to make it possible to run complex models in minutes rather than days, with secure access to a common pool of data.'

As populations rise in our towns and cities, the demand for new and improved infrastructure - housing, roads, electricity, clean water etc. – rises with them. Now, more than ever, we need to analyse every effect on the environment associated with building new infrastructure, including the impact on the quality of the air that we breathe.

Computer modelling now forms a vital element in infrastructure design but at present it is extremely difficult to consider all aspects of infrastructure and how they impact on one another at scale. There has been no central resource available for researchers to share knowledge and analysis expertise to build new analytics tools… until now.

Says Marion Samler, the DAFNI Partnerships Manager: 'We’re providing a facility where organisations can lodge their data on a secure platform, upload their models, and link to the data. This enables large scale analysis and, crucially, provides the visualisation tools so that results can be shared in a meaningful way. Keeping all the relevant data in one place with the models makes it easier to take all potential impacts into consideration.

'For instance, we are working with Urban Observatories run by researchers in six major UK cities. They are collecting all sorts of city and environment-related data from sensors placed in strategic spots – so they have information on things like local air quality, weather conditions, traffic flows and energy use.'

The DAFNI version 1.0 ‘system of systems’ is currently being trialled by universities, government and industry. DAFNI is funded by the UK Collaboratorium for Research on Infrastructure and Cities (UKCRIC), a collaboration of 12 University partners, with Oxford University as the lead partner. It is hosted by STFC’s Scientific Computing Department in Building R89 at RAL.