Saturday, 13 February 2016

My unconventional research career - from industry to academia and back again

I was recently asked to give a seminar at the Australian Meteorological and Oceanographic Society Early Career Researcher Forum on my research career to date. I got quite a good reception and thought that I would share my talk here, in the event that others might also gain something from my previous experiences.

Good afternoon,

Firstly I'd like to thank the organisers for giving me this opportunity to reflect back on my research career to date, and hopefully share some useful knowledge and experiences with you all.

Like the other participants I've been asked to give a talk on my career and to provide some advice to you all. Perhaps, the best advice I can give you, however, is not to listen to anyone's advice! Seriously, the most important thing you can do is to first try to figure out what it is you want out of your career and life in general. Next talk to as many people about their own careers as possible, not in search of career advice per se, but instead in search of career examples from which you'd like to replicate certain elements. The research sector globally is going through many changes at the moment, so any advice you're likely to receive may be outdated, and in any case filtered by the experiences of the people who are giving you that advice. So today I'm going to give you an overview of my somewhat non-standard research career, from industry to academia and back again. At the end of my talk I'll also provide some concrete tips on how to make the transition from academia to industry if that's what you're after.

I've worked in various sectors in industry and academia over the past 12 years, primarily in the fields of numerical simulation and data mining. My journey started with my undergraduate degrees in mechanical engineering and finance at the University of Western Australia. In 2001, at the end of my third year I undertook vacation employment with BHPBilliton in Port Headland in the north-west of Australia, working in the iron ore ship-loading division. At the end of the following year I did vacation employment with the Defence Science and Technology Organisation (DSTO) in Adelaide, undertaking computational fluid dynamics analysis of the next generation of missile propulsion systems. I found the science both challenging and interesting, and I also had an excellent mentor. After this summer vacation work period, I returned to university to complete my degree. I applied for a graduate position at DSTO, of which I was the preferred candidate, but at the last minute the funding was pulled from the position.

Following this I then applied for and attained a continuing graduate industrial research position at General Motors Holden Innovation. I worked there from 2003 to 2006, where I was essentially a liaison between production engineering and academia on developing virtual engineering solutions for the engineering design processes. For example, using numerical optimisation to minimise weight in suspension components as opposed to running many physical tests [1]. Despite the pay and conditions being good, I personally found my position not all that fulfilling, as I was an intermediary, and the “real” work was being done by our collaborators in academia and at the Victorian Partnership of Advanced Computing. I unsuccessfully tried to get back into DSTO on several occasions, and realised that completing a PhD would help me get back into that organisation.

During my time at Holden I met many academics as part of my work. I decided to do a PhD at the mechanical engineering department at the University of Melbourne (with co-supervision from Monash University) in the numerical simulation and reduced order modelling of aerofoil flows, applicable to the improved energy efficiency of wind turbines and aircraft. It was actually a double badged degree with the Université de Poitiers in France, where I worked for almost a year and a half with some fantastic people. During this project I also worked with some incredible people at Stanford University, the Universidad Politecnica de Madrid, and the University of Manchester. As an added bonus I also met my now wife at the Poitiers swimming pool, but that's another story! I completed my PhD in 2010, produced four journal papers, and also won a prize for my thesis [2].

Upon completion of my PhD I was offered 6 different post-doc positions, both here in Australia and overseas. One of the options was a 3-year contract role at the CSIRO Marine at Atmospheric Research group working on developing the next generation of climate models, to better inform policy makers on issues pertaining to climate variability and change. I was sold on the role immediately after meeting the impressive people with whom I would be working. It is interesting to note, that one of the other options was a continuing role at DSTO, the organisation that I had been trying to get back into for the previous 7 years. I worked very hard over a long period to get back into DSTO, however, at the point that the opportunity was offered to me, I didn't want it any more.

I worked at CSIRO from 2010 to 2013, during which time I learnt an immense amount from my co-workers. I solved the long-standing and important problem of resolution dependence in atmospheric and oceanic climate models, via the development of stochastic subgrid turbulence models [3]. As my contract at CSIRO was coming to an end, it became clear that even though I wanted to stay and the organisation wanted to keep me, budget cuts meant that this would not be possible unless I brought in some form of external funding. I unsuccessfully tried to get additional funding via the Discovery Early Career Researcher Award (of which CSIRO is no longer an eligible institution). After this I then gave seminars at various geophysical research groups in an attempt to identify any potential opportunities for co-funding between a partner university and the CSIRO for me to continue the research. Unfortunately all of the university groups I approached had their own funding problems and were struggling to keep hold of their own researchers at the time. I was reluctant to search for positions overseas, since my wife had already relocated for me once, made new friends here in Australia, and had also started making inroads into her own successful career as an educator.

I then approached my former PhD co-supervisor from the mechanical engineering department at Monash University, who was awarded Australian Research Council (ARC) funding to undertake the world's largest numerical simulation of an adverse pressure gradient turbulent boundary layer [4]. We organised a 3-year arrangement. I would be a Monash employee working on this extensive and challenging ARC project for half of my time. CSIRO would then pay Monash for me to continue the subgrid turbulence model development for climate simulations in the other half of my time. In addition, during the last two years of the arrangement I lectured a final year elective unit at Monash University on advanced aerodynamics and turbulence.

This arrangement was very productive, however, it took its toll. After two years of working every night and every weekend, and unsuccessfully applying for various lecturing positions, I was exhausted, burnt out, and felt that all my efforts in attaining a continuing academic role were futile. I have always prided myself on working on difficult and important problems, but this meant that my publication rate was less than some of my competitors. I have averaged two first author journal papers per year throughout my career, but as I've come to realise, to succeed in academia two papers a year is not going to cut it. What's more I realised that I'm not willing to work the way academia demands of me, in terms of many studies solving incremental problems, as opposed to fewer studies solving big, risky, and difficult questions. Another factor in my decision, was that my wife and I wanted to start a family soon. I felt that we needed some more stability, and I needed a better work/life balance. This made me search for other career options, and in doing so I realised that there are opportunities for challenging, interesting and meaningful work outside of academia.

For those of you toying with the idea of perhaps working in industry you might get something out of the steps I took over the past year. The first thing I did was to put a blog together communicating my previous research in lay terms, to serve as a portfolio of work to share with potential employers. I also created a website, youTube channel, Google+ page, twitter feed, and LinkedIn profile to build upon my online presence. I started attending a couple of after-hours industry based meetups in the field of data science, and made sure that after the seminars I got the contact details of at least one industry representative. I also started collecting online job descriptions of positions that I thought I might be interested in, and soon realised that many of the roles were looking for similar skills and experiences. Once I determined the skills I needed to develop I put two resumes together: one representing my current skill set; and another target resume that I would like to attain. I then made plans to address whatever short comings I deemed that I had. In addition, over the past year I have been doing some advising/consulting for start-up eHealth company, CurveTomorrow, using machine learning methods to build predictive models linking patient details, initial diagnoses and initial procedures to the probability of the patients returning to the hospital with future complications. This experience served as another example of my ability to quickly determine the important factors in a new field, and apply my mathematical and numerical skills to solve their problems.

As a result of these efforts I had job offers to work in the energy sector, management consulting and also for the global macro fund manager, AE Capital. I accepted the AE Capital role and have been working for them since the beginning of the year. The reason I chose AE Capital above my other options, was that I felt that I could bring some new ideas to their business, contribute in a meaningful way, and felt that I could learn much from their current team. I still publish research with my academic collaborators to contribute to the wider community where I can.

In summary, as I mentioned in the beginning, the best advice I can give you is first to figure out what you want, and talk to others in search of career examples to draw upon, as opposed to specific career advice. Above all try to work on things you love, and be good at it, so that you may have an opportunity to continue to work on the things you love in the future.

Post Script:
As of July 2017 the opportunity came up for me to return to CSIRO in a continuing position working on a long term decadal prediction project. Working in finance was interesting, I learnt a lot and met some great people. However, climate science is my passion and something I want to dedicate the rest of my life toward.

[1] Hermann, P., Ting, T. & Kitsios, V., 2005, Development of a fatigue analysis tool chain for automotive structural applications, Society of Automotive Engineering Transactions, Vol. 114, no 6, pp 522-530.[link]
[2] Kitsios, V., 2010, Recovery of fluid mechanical modes in unsteady separated flows, PhD Thesis, The University of Melbourne. [link]
[3] Kitsios, V., Frederiksen, J.S. & Zidikheri, M.J., 2015, Theoretical comparison of subgrid turbulence in the atmosphere and ocean, Nonlinear Processes in Geophysics Discussion, Vol. 2, pp 1675-1704. [link]
[4] Kitsios, V., Atkinson, C., Sillero, J.A., Borrell, G., Gungor, A.G., Jiménez, J. & Soria, J., 2016, Direct numerical simulation of a self-similar adverse pressure gradient turbulent boundary layer, International Journal of Heat and Fluid Flow, under review.

No comments:

Post a Comment