Find research collaborations for programming and publications

Question

Hi,

Let's say that I work on a public research organisation, as researcher with PhD, background in biophysical sciences and lots of PostDoc experience and that I need to get some number of scientific publications in high-impact journals each year in order to get extensions of our contract.

Let's say that I am really competent in some new technology like GPU programming, I mean, better than average.

Therefore one of my objectives is to find other research groups, who have developed some important biophysical model and computer program, but who have no experience in GPU programming, and to suggest to them to start a collaboration, so we get finally the method running on GPU's. This is great because later we can get publications and all of us will be happy.

My question here is what approaches do you suggest if any in order to:

1. find these research groups interested in possible collaborations
2. if not, how could you tell a group the interest of getting their method running on GPU

What I have tried now is:

1. going to conferences in your field. But now, for personal and health reasons, I cannot move from my place

2. find publications which could be related with these groups, and try to contact them, but this approach rarely works

So, these are my questions. I think that this is neither vague nor not programming related.

Thanks in advance for your help

Answer 1

Join professional societies and read journals in your area of specialty. Read the public areas of collaboration web sites. Attend conferences (on your own dime, if need be). Talk to people. Start working on a (any!) problem in your "spare" time---something related to recent papers even if not cutting edge, just to show that you have something to offer. Schmooze people at the local college or university, get to know them, and offer to give a colloquium (most departments have trouble filling the calendar and will jump on any offer they are sure won't embarrass them); then make the talk good. Communicate with people: in person is good, but email is acceptable.

When you are ready, you will approach a member of the collaboration directly. Ready means that they already know who you are (and don't think you are a crank!).

It's best if you can approach a senior person, but getting the sponsorship of any full member (professor, lab staffer, PI from a private sector organization; not a post doc or grad student) is a chance. Be prepared to tell them what you can do for them and back it up. Be prepared to sign on for some unglamorous grunt work, too.

The "don't think you are a crank" thing is a real issue, BTW. Without the cachet of a known institution backing you up, you are going to have to prove that you're up to the job every step of the way. But you can get some help from your adviser and your colleagues from times past. Let them know that you are in the market, and ask them if it is OK to drop their names. Ask a few if they'd be willing to write you letters of recommendation when the time comes. Ask them to introduce you to people. The good ol' boy thing isn't there as some evil plot to keep out the hoi polloi (really it is a defense mechanism), but it is there. Use it to your best advantage.

But, probably don't ask on Stack Overflow... as it is not really programming related.