Research And Development

To qualify as ResearchAndDevelopment, a project should generate either

patents, or
papers acceptable for publication in a peer-reviewed journal.

Often the term is abused, and is used to describe straight development.

I don't see that. It should generate knowledge, but there's no particular reason why it should be required to be made public in either of those ways.

Concur. Many times I have been involved in R&D projects that simply expanded known technologies into unknown applications. Does this not qualify as research? It is still ground-breaking work that no-one else has attempted. True, it is mostly engineering, but the research element can't be dismissed simply on the grounds that it doesn't generate a patent or peer reviewed white paper.

No, that doesn't qualify as research. All software development is ground-breaking work in this sense, unless you're just copying someone else's code verbatim. So being R&D should mean something more than that, like discovering a new technique, and then taking time out to study it before writing about it. I'll concede, though, that if you intend but fail to publish it for whatever reason (e.g. get it rejected, or resign before you complete the research), it might still qualify.

Whatever it's called wouldn't matter, except that some governments give tax breaks to companies doing "Research and Development". Calling straight development R&D in order to claim those tax breaks is fraud, pure and simple. -- DonaldFisk

[I don't think dictionary.com agrees with you: http://dictionary.reference.com/search?q=research

And surely the definition of research that matters for tax breaks is whatever the regulations governing the tax break define. I imagine that depends on the country, and the scheme.]

Hmm. Hey, look - I've worked on plenty of medical products that required research before and during development. One Nuclear Medicine project in particular was being supported by a handful of full time Nuc Med scientists and physicists and also a part time mathematician. Of course there was the usual mix of electronics, software, and other disciplines involved as well. The point is that we were on the bleeding edge of a technology that pushed the envelope of the science itself. We didn't know exactly what we were making until we made it and had a chance to try it out. The science actually followed the technology along part of that curve.

Another medical product I worked on was an apheresis blood separator that was originally intended to extract platelets from healthy donors for use in treating post operative leukemics, hemophiliacs, and other special needs patients. The medical research boys discovered that the magical machine could be used for other purposes, like actual real time treatment of certain blood-borne pathogens and the filtration of particulate toxins. This research into using the machine for actual treatment was ongoing long after I left the project.

Now, does this kinda stuff not qualify as research? Some of the technology involved - like the contrast enhancement algorithms of the Nuc Med box - has been around for an eternity. However, the actual medical application required a whole new realm of science be developed to properly use that technology and then refine the tech to meet the medical needs. This wasn't just a matter of plug and chug with numbers. And what about the treatment via apheresis? Holy mackerel, there, Saphire - we're talking about totally new treatment protocols here!

Okay, so these examples did generate scientific publications. I have also helped develop products that broke new ground but didn't necessarily generate excitement in the scientific community. Still, the research had to be done to know what a product, technique, or even a whole technology is good for. Your definition of research being limited to scientific publication is way too restrictive to be of any use to most of us here, methinks.

-- MartySchrader

I've seen a lot of boring papers, but they were published and they were research. Also, a lot of research results are negative, and people are less inclined to publish that. But we've shifted subject off of R&D and onto scientific research. -- DonaldFisk

There are all sorts of meanings for "research".Market research is going out and asking people questions, and then analysing their answers. It's still research by anyone's definition. But it's not what I meant here. What I meant was ground-breaking work that's never been done before, done usually for its long term value.

That's a useful concept which shouldn't be diluted by things which lack any element of originality, done for a company's short-term benefit. -- DonaldFisk

You seemed to be defining research for the purposes of contrasting it with "fraudulent research" when getting tax breaks. Although your point of view is interesting, isn't it the one intended by the authors of the tax break legislation that matters for that? Anyone have some examples of how they define "research"?

In other words, why is your definition the right one of the purpose of "R&D credits"?

In the UK, it's http://www.inlandrevenue.gov.uk/r&d/dti_guidelines.htm This does not require the research to be published, but it's clear to me that it has to be acceptable for publication. -- DonaldFisk

Thanks for the link. I don't read it that way, though. Note, also, that that document doesn't seem concerned with the difference between R and D - but between R&D and non-R&D.

I was talking only about R&D. It's one thing.

''From the top of the page: "Often the term is abused, and is used to describe straight development." Hence, I thought you were differentiating between the two.

Research is a lot of different things. The Inland Revenue link might not say that R&D has to be acceptable for publication in so many words, but the sort of work it describes certainly will be patentable or publishable, in some peer-reviewed journal or other, should the authors decide to submit. -- DonaldFisk

We'll just have to agree to disagree on that one. I don't think publishing a lot of stuff covered by the Inland Revenue definition would be at all interesting or useful. You can peer-review anything, after all.

In many cases, a company will choose to retain its intellectual property as a confidential "trade secret" rather than patent it, which makes it public. I think the claim that patents or papers are required is absurd. If I compose a song but don't publish it or perform it, have I not really composed? I can't be credited since that's a public affair, but it'd still by my song.

Patents may be public, but until they expire the patent owner can either request a royalty or simply refuse to license the invention. But to answer your point, if the work's worthy of publication in a peer-reviewed research journal, then it's research. But there's no way of knowing for sure until it is published, and I doubt if that sort of thing goes on much outside the military. -- DonaldFisk

I work for a company that has made significant advances over the past few years in our field (e-learning applications). We do not publish our advances and yet we still claim credits for them. As one of the people that participated in the credit claiming process, I can tell you that everything we claimed credits for was indeed researched and considerable effort was put into coming up with the concepts involved. Research is research whether or not you decide to make it public.

I have already clarified that I would count it as R&D if it's acceptable for publication in a peer reviewed journal. If you can honestly say that your work is, then it's R&D for tax purposes in the UK, and also by my (revised) definition. I would, however, urge you to publish your research as soon as it no longer needs to be kept a trade secret. From a purely political standpoint I don't like the idea of tax credits being given with nothing ever being returned to the community. -- DonaldFisk

There's a technology continuum from pure research to product development and introduction into a market (and its probably not just one-dimensional). At various points on the continuum different practices and cultures are typical. One model for this progression is "Technology Readiness Levels"; a web search will turn up numerous references in military and aerospace domains. While the model is hardware-oriented, it is often used for software too, for example, avionics or other onboard/embedded software, and information systems. At lower levels key "artifacts" are reports, table-top experiments, and so on. At the higher levels, prototypes, integrated sub-scale systems, demonstration of product-level capabilities, and finally proven functionality as a full-fledged product. I think rather than requiring work to be published in a peer-journal it is sufficient that the work be in a state where it can effectively move forward toward the goal of a technology deployed in the marketplace. In practice that means that the knowledge has been captured in technical reports of some kind and working devices/software/systems. Requiring publication or patents strikes me as too stringent - both require substantial additional investment of resources not directly related to reaching the technology goal. In fact, some companies limit the number of patents they apply for to keep the associated costs down - but that doesn't mean they aren't doing R&D.

There are big cultural differences between R&D and straight (product) development. R&D is optimized for obtaining useful knowledge and demonstrating new capabilities. Product development is optimized for cost-effectiveness and customer satisfaction. These different foci yield very different ideas of what quality means, and consequently the processes and cultures are different. I have big challenges with my teams navigating that divide, but there are huge payoffs for any team that can - and for their organization.

From a project standpoint, there is one big difference between R&D and straight development. R&D is very hard to estimate unless the technology increments are very small. So don't put R&D work on or anywhere near your project's critical path. From a project manager's standpoint, R&D milestones should be a pre-requisite for project initiation, but the disadvantage of this is that it results in long product cycles. An alternative approach that is better strategy for the overall business is to plan technology infusion points at various points along the project where new technologies can be incorporated - but always with ready alternatives in place that can be used if needed. This can be more costly, but in the end better for the business since the product can ship earlier with at least some new technology that customers find valuable. This strategy is especially useful for multi-generation product families since if a technology misses its infusion point for product generation 3, it can still make it into generation 4 so the investment is not lost.

Finally, regarding defining R&D based on taxation: lots and lots of R&D is done by academic, non-profit, and government institutions not subject to taxation in the same way many businesses and individuals are. Any tax-based definition of R&D starts to get really complicated and is probably best understood with a tax lawyer in the loop if the answer really matters in any specific circumstance. But the broader definition of R&D should be based on what is done, not the environment in which it is done.

-- DaveVanBuren

Refactoring call for here after we hash this out a little more.

I have to spam intention but rather suspicious of the manipulation of content.