In march2012, a group of like-minded software developers gathered at Oxford University in the UK to participate in their so-called "collaborations workshop" seminar. They have a common profession - developing software to support scientific research, but their positions are different, and they do not have a clear career path Participants created a term to describe their work path: Research Software Engineer (RSE)
Ten years later, RSE Association Rose in Britain, continental Europe, Australia and the United States. In the UK, at least 31 universities have their own RSE groups. Paul Richmond, head of the RSE group at the University of Sheffield and former chairman of the British research software engineering society, said that this shows that the importance of this profession is increasing. Chris woolston, a reporter from nature, interviewed Richmond and learned about his life as RSE, the role of software in research enterprises and the status of this field at the tenth anniversary.
Nature: what does RSE do
Richmond: fundamentally, RSE supports scientific research by developing software. They usually don't have their own research problems -- they develop computer tools to help others do cool things. They may add functionality to existing software, clear errors, or build something from scratch. But they don't just sit in front of a computer and write code. They must be good communicators and be able to integrate themselves into a team.
Nature: what kind of projects are they engaged in
Richmond: almost all scientific fields run on software, so RSE can find itself doing almost anything. In my career, I have worked in cancer cell imaging and pedestrian traffic modeling software. As a postdoctoral student, I am engaged in computational neuroscience. I don't know much about these specific research areas, so I work closely with oncologists or neuroscientists or someone else to develop the required software.
Nature: why do so many universities support their RSE teams
Richmond: some high weight researchers at the top of the academic echelon have the ability to hire their own RSE. The Engineer may be dedicated to maintaining a single piece of software that has existed for 10 or 20 years. However, most research groups only occasionally need -- or can afford -- an RSE. If their university has an RSE team, they can hire an in-house engineer, one day a week, or once a month, or anything they need. In this way, the RSE team is like a core facility. Universities try to ensure that the team has a stable workflow, but this is usually not a problem -- there is no shortage of projects to do.
Nature: what else does RSE do
Richmond: a large part of RSE's work is to raise awareness of the importance of high-quality software. RSE may train a postdoctoral or graduate student to develop their own software. Or they can hold a seminar on good software practices. In theory, training 50 people may be more influential than working on a project. In practice, because they are busy supporting research, RSE often find it difficult to find time for teaching, guidance and publicity.
Nature: does the principal investigator (PI) understand the demand for RSE
Richmond: it's complicated. In the past, researchers were not always motivated to use or create good software. But that is changing. Many journals now require authors to publish codes that must be fair: searchable, accessible, interoperable, and reproducible. The last word is very important: good software is an important part of the study of reproducibility. RSE explained to the chief scientist that they needed reliable code so that they would not have to withdraw their papers after six months.
Nature: who should consider working in RSE
Richmond: many RSE started as doctoral or postdoctoral students, working in software to support their own projects. They realize that they prefer this part of the work to the actual research work. Rses certainly have the skills to work in industry, but they thrive in the cutting-edge scientific environment of academia.
Most RSE have a Ph.D. - I have a Ph.D. in computer graphics - but this is not necessarily a requirement. Some rses end up in tenure; I was recently promoted to professor. Many others work as laboratory technicians or service personnel. I would encourage any experienced developer interested in research to consider RSE as a career. I also hope to see more people from underrepresented groups join this field. We need more diversity to develop.
Nature: do you have any suggestions for those who hope to become RSE
Richmond: you can try to develop an open source software. If possible, do some training in a cooperative environment. If you have any questions, you can talk to the RSE who is working. Consider joining an association. The British research software engineering association is happy to provide people with advice on how to enter this field or how to stand out in job applications. People in the United States can contact the American Society of research software engineers.
If you are a doctoral or postdoctoral student, give yourself a challenge: try to convince your supervisor or PI that they really need to accept good software technology. If you can change their minds, it is a good indication that you have the passion and motivation to succeed.
Nature: what do you think about the development of this industry in the next 10 years
Richmond: I hope to see that RSE is equal in the academic environment. Software runs through the whole research process, but professors often get most of the recognition and reputation. The influence of a software is as big as that of some research papers, and some are even greater. If rses can get the recognition and reward they deserve, then this career path will be more obvious and attractive.