Scientists of today use data processing software to make discoveries, but it takes the right software to crunch the masses of data and present it to be analysed.
One challenge faced by scientists is that every researcher who approaches a published dataset has to analyse it from scratch, which is a huge drain on time and resource.
This is where Dr Ed Bennett from Swansea University comes in.
Ed, a Senior Research Software Engineer, is helping transform the way data is used and shared by scientists, one research community at a time, with the support of a Science and Technology Facilities Council (STFC) fellowship. The newly-awarded Research Software Engineer (RSE) fellow will help researchers make scientific discoveries in a shorter period of time and enable high-quality research.
This RSE fellowship call (run jointly with Engineering and Physical Sciences Research Council) recognises the vital role of software in research and supports ambassadors with potential to be a future leader in the research software community.
Ed plans to improve software used by research communities, to support and train other researchers and to conduct his own research.
Though research software engineering is one of the newer engineering careers, they are vital members of computational research teams. These engineers ensure the teams have access to the skills required to develop, support and evolve software to tackle the research challenges they face.
“In the last decade, there has been a realisation that research is only as good as the software it uses and if there are mistakes in the software, there will be mistakes in research results. It is my aim to promote the role of research software engineering so that it will sit alongside more traditional academic and research roles.”
One of the research communities Ed will be supporting are Lattice Gauge Theorists.
“The Lattice Gauge Theory studies interactions, including the strong nuclear force that holds protons and neutrons together and also other strong interactions that we haven’t yet seen in nature. To understand the interactions we need to use numerical approximations, with support from supercomputers.”
Researchers use powerful supercomputers to generate the complex data in a shorter period of time. However, it still takes a while to produce enough information to draw conclusions. The challenge is more acute when the research field is very collaborative.
One issue that many researchers encounter, including Ed, involves updating or extending an analysis with unsupportive software.
“One frustration that I encounter in my research is when I go to update or extend an analysis I did a while before, but I have forgotten exactly how it was done. This means I either have to re-do the entire analysis, or present results that are not consistent.”
Ed aims to work with the Lattice Gauge Theory community to develop a new, fully reproducible data analysis software. The cutting-edge tool would allow researchers to re-run the software used to create the tables and figures in a publication, and get output identical to the published version. It could help researchers make scientific discoveries faster than before.
Another challenge Ed hopes to solve involves improved access to other researchers’ analysis.
“Researchers in Lattice Gauge Theory will often share only the results of their analysis. This means if other researchers want to check the analysis, they will need to repeat it from scratch.
“This can be a huge effort.”
Ed aims to work with the community to define a set of common standards and practices, and develop tools to make it easier for researchers to follow. Researchers would then be able to share project results and techniques more easily with the community, so that they can be reproduced by others without the guesswork.
“My hope is that I’ll be able to drive research communities towards workflows that are efficient, both in terms of how long it takes the computer to calculate them and how long the researcher has to spent pressing buttons.”
With support from the RSE fellowship, Ed will also collaborate with Aberystwyth and Cardiff universities to improve the performance of the software they are using to analyse solar and gravitational wave events, respectively.
“Being able to help others tackle problems that they couldn’t otherwise have done gives me a huge thrill. For example, meeting someone who was waiting days for an analysis to complete on their own laptop and guiding them through running it in minutes on a supercomputer, really makes me feel like I’ve been able to make a difference.”
From a young age, Ed was interested in physics and how things work, and how people can work together to further understanding.
“I’ve been attached to computers since the first time I touched one. The fact that you can get results not only faster than by hand, but also more consistently, is as close to magic as I’ve come.
“The fact that others don’t always see computers that way is one of the things that really drives me, to help people get more effective use of the technology that is available.”
The RSE fellowship was first introduced in 2015 by the Engineering and Physical Sciences Research Council.
Last updated: 27 May 2021