Academic Awareness in Industry

02/19/2011 by Sylvain Hallé

I may have been a researcher for only a couple of years, I already can't keep track of the number of meetings, conference panels, plenary sessions and corridor discussions whose topic revolved over what seems a perpetual question: what does industry want?

As members of academia, we are routinely enticed to produce results with potential for practical applications. For better or for worse, the situation has come to the point where we see the reuse of our results by industry as the sole purpose of our work and its ultimate validation. If this doesn't work out, the heart of the problem, it always seems, is that we academic researchers just don't try hard enough to reach out. And so at least once a year since I'm doing research, I have been caught without even looking for it into one of these discussions where academics nervously speculate on a community strategy to grasp the attention of those "industry people".

It was in this deeply rooted mindset that I started reading a paper in the latest Communications of the ACM, where a seemingly trivial remark caught my eye:

"Although pockets of interesting academic research exist […], as an industry we've found it difficult to translate academic work into practical value. Sharing of knowledge and tools from academia has been limited, as companies need first to pay for access to many of the papers and then translate the formal structure of those papers into something that makes sense to them." (p. 46, emphasis added)

As an academic researcher, I was deeply surprised, to say the least, of such a comment. Here apparently stand the Two Great Barriers that prevent knowledge transfer to industry:

1. "Companies need first to pay for access to many of the papers".

Apparently the practice of Googling is not so widespread after all. Virtually all authors keep PDFs of their works on their personal web sites, and if not CiteSeer might even keep a copy in its cache. I have rarely failed to find a free version of any paper I needed. But even then, buying a Springer LNCS paper costs $25. IEEE journal papers cost $19. ACM charges $15, and even offers view-only access for $3 per 24-hour period. Heck, if you're so short on money, you can visit any university library, view the paper (most universities can afford electronic journal subscriptions) and print or even screen-scrape it.

Yet the author of the previous quote works for eBay, the company that boasted $1.7 billion in profits in 2010. If these folks refrain from paying 25 dollars to read a paper that might solve a problem they have, academic research is doomed no matter what we try. Forget about outreach research grants, collaborative partnership programs and technology transfer initiatives if the major hurdle is the cost of a yearly journal subscription. We'd all better close shop and find something else to do for a living. Unless I miss something important, the cost excuse does not withstand the most superficial examination.

When our current prime minister first got elected, author Yann Martel started mailing him one book every other week, in part to highlight the little care given to culture in his new government. Should we start mailing research papers to industry?

2. Companies must "translate the formal structure of those papers into something that makes sense to them".

Says a Chinese proverb (sometimes attributed to Confucius): "Man who stand on hill with mouth open will wait long time for roast duck to drop in."

Alas, research papers are not diff files that any company can merely pick up and apply to patch its latest code branch. Such a comment makes industry appear like a fledgling waiting for academia to beak-feed its results into immediately consumable niblets in order to be worthy of any attention. Besides, if my work is so specialized that it does not require any adaptation for some business to use it, then I am pretty much working for them and I sure hope they are paying me for producing such tailor-made stuff.

By definition, academic work is generally not concretely usable without a modicum of effort. By definition, R&D implies some form of passing the baton between theory and practice that involves work on both sides. The fact that some people in the industry see this work as a liability rather than an investment is very discouraging for anybody looking forward to fruitful academic-industrial collaborations of any kind.

The next time I find myself in the middle of another academic neurosis, I'll try to remember the advice a colleague gave to me this week: the best way to transfer your knowledge into a company is to start your own. Businesses are much more receptive to a successful startup competitor than to the same ideas floating around the web in a free PDF.

1. J. Harty. (2011). Finding Usability Bugs with Automated Tests. Communications of the ACM, Feb. 2011, 54 (2), 44-49.

comments powered by Disqus