Historically, the external funding of the FSU Department of Computer Science has been a mix of small (tens of thousands of dollars) and medium (hundreds of thousands of dollars) contracts and grants. Last year, the eleven tenure-track faculty spent about half a million dollas on sponsored research activity. Over the past few years, it seems that at any instant about half the faculty have some form of external research support. Some of these have been grants for basic research, but the trend has been toward applied research and development contracts.
The primary expenditures in Computer Science contracts and grants are generally for personnel. Most researchers do not need equipment that compares in cost to particle accelerators and NMR equipment, or specially fitted laboratory facilities. People doing software-oriented research and theory typically can get by with a modest network of desktop workstations, working in ordinary office space.
There are exceptions, of course. There are some hardware-oriented projects, that involve the design and custom building of new computer or data communications gear. There are also interdisciplinary research projects in Computational Science and Engineering, which may require massive computer resources to solve computationally complex applications. Funding for the large-scale computers to support the latter is generally via separate infrastructure grants, similar to those available to people who need lab equipment in other disciplines. The NSF hands out a few such multi-million dollar infrastructure grants each year.
Over the past few years, funding for research and development in Computer Science has evolved. One trend has been toward fewer medium-sized (hundreds of thousands of dollars) grants. Most government spending is going to large-budget projects, involving one of the elite established larger research departments or a consortium of strong departments. The rationale heard from funding officers is that with shrinking research budgets the way to maximize benefits is to concentrate the money on a few strong prospects. Smaller grants, such as to provide travel and summer support for a single faculty member, are still available for basic research but funding agencies tend to be stingy with money for research assistants, especially at schools that are not already viewed as being of the first tier in Computer Science. It is easier to obtain funding for research assistants if they are for applied research and development projects.
The biggest sources of funding for research in CS are:
The DOD spends money for basic research in CS primarily through the following well-known agencies:
These all tend to focus increasingly on projects with specific anticipated useful products, or targeted at solving problems of specific interest for national defense. The areas of interest change over the years, so to obtain continued funding generally requires the ability to change areas of research. In addition to the agencies listed above, defense funding is available through subcontracts issued by various government laboratories and defense contractors. Such contracts are even more tightly focussed on specific short term results.
Defense spending has been shrinking over the past few years, including funded research at universities. With cutbacks in major weapon systems development projects, the large defense contractors have increasingly come into competition with universities for the remaining sources of research and development funding. Cutback in funding for the Federal laboratories have also had a similar effect.
These are not the only funding sources. Funding for interdisciplinary work involving CS is available through other federal agencies, such as the Department of Energy, NASA, and the NIH. There is also a small amount of funding for academic research available from industry, but this is limited due to concerns about trade secrets and intellectual property rights. What is said above about targeting specific areas of interest also applies to these sponsors.
It is probably safe to to say that the more applied the CS research project is, the easier it is to find funding, but the smaller is the fraction of the funding that can be applied to publishable academic research. Most of the money spent on CS research is is focussed on applications of computing to a specific area, rather than core CS research. Many of these more application-oriented contracts either explicitly limit publication, implicitly discourage it by requiring delivery of software and special reports of a form that cannot be published, or end up producing publications that are in the area of the application, rather than in CS.