A few weeks ago I wrote about Trickle Down Technonomics, or more simply: how large investments in prior research can and do still affect us in the modern day. Since then, an article from the author of a forthcoming book on Bell Labs wrote an intriguing piece in the New York Times. I highly recommend checking it out. And while I haven’t changed my mind about the necessity of high-level research and investment, there was a glaring problem with the last post and my focus on high level: it’s not realistic these days.
I’m not sure about the socio-political feed-ins that allowed raw research investment back then, aside from the specific example of AT&T’s glaring monopoly in the case of Bell Labs. What I do know is that research for the sake of research in a for-profit company seems to makes people (managers) more squeamish these days. Perhaps it was the cold war that drove this thinking and investment? Or other wars? Or maybe it was the populace being more enamored with technology and science because it wasn’t quite as prevalent in their lives? Perhaps if today we did not have the ability to communicate half way across the world using a device in our pocket, maybe we would all dream of one day being able to do so? Why this doesn’t extend these days into the next steps, I don’t know. But what I can see is that we don’t fund large labs these days, often due to concerns about Return on Investment (ROI) and other stalwart units of measure from the MBA Playbook.
But is this all really true? Are there no places that are forward looking in their technological outlook? I’m sure the academic-based engineers on our site would beg to differ. In fact, the university lab is possibly the best example of cutting edge research, simply because they are tied to grants and paper pubishing versus product deadlines. They are able to look forward not 18 months to the next product launch, but 10 years to where the technology needs to be. It seems there is continuing effort to push more of this research onto university labs. There are research collaborations with industry, sharing costs and any output of work/publishingone example of how university labs are expected to take on the burden of seeking innovative new ways to do things. Another offloading of the “R” in “R&D” (Research and Development for the lingo-challenged) is technology transfer offices at universities. These instead take research that was not directed towards any particular purpose and then attempt to license the idea or spin off a company into a real world product. These are attractive because they benefit the university which officially owns the intellectual property (IP) and will continue to profit from the revenues of any licensing deals.
In the private sector, we do see pockets of forward-looking thinking. Often these instances pop up at companies that are doing exceptionally well monetarily and have talent to spare; managers tasked with innovation help too, as they are important in shielding top thinkers from the everyday drag of bug fixes and fire fighting. The most obvious example here is Google and their X Labs. Recently, there was an unofficial announcement about the future release of glasses which will be able to interface to mobile phones and display data in an overlay in a pair of glasses or sunglasses. However, their not-so-successful cousin Yahoo! also has a labs group, one that even went so far as to predict the outcome of the 2012 elections (not quite the same “wow” factor).
If we ignore some of these institutions that persist in a post Bell Labs world, what do we have left? An email exchange I had with element14 author and EE friend CJ Gervasi highlights some of the potential benefits of a research group of the future. CJ writes:
I am interested in the engineering trickle-down idea more because it does seem like we’re seeing the decrentralization of R&D, and the decentralization of the things due to the internet, and I want to understand if that’s what you’re saying and what you think it means.
CJ wrote about the future of R&D and the distributed model of open source. And that’s a point I completely agree with. The open source movement is a perfect example of how we see innovation on a small level that then moves back up the chain, into the source of either a software or hardware project. I have experience with the Open Source Hardware (OSHW) side of things, mostly due to my exposure to it on my electronics podcast. I regularly see projects that blow my mind, most often because of the ability of the project creators to do so much with so little, often building on others’ work. However, this also makes me feel that I should define innovation, at least as I see it.
In my response to CJ, I told him about a situation I was in:
I got myself in a little trouble at a past company when I sat with a VIP and told her, “I don’t think we innovate enough”. She then rattled off stuff we had done, which truly was innovative. The difference was that was on a small (subsystem) level and was iterative from past products. We also innovated on a process level, both in the workarounds my sustaining group had to do on a regular basis and the everyday manufacturing changes we needed to make products better. However, we didn’t have any system level innovation. Perhaps because they’ve already got it all figured out? Not likely. Competitors were creeping in from the wings because of the same reason any established player in a mature industry falters: giving customers what they ask for, instead of what they’ll ask for in 10 years. Too much fear to do something like that. But I digress. I think that the system level innovation goes away, and possibly some of the pieces (circuits) innovation as well. For example, a GPS receiver is a “piece innovation” whereas the entire GPS network with all its complexities is a system innovation.
What I was trying to get at is when someone says, “Innovation”, it can mean a lot of different things. And if the top level system integration isn’t there, industries are doomed to continue refining ideas and never making leaps to the next big thing. Some of this is happening today in the open source movement.
One example is MakerBot, a 3D printing company that uses extruded plastic to create physical models of a 3D CAD drawing. One of the founders, Bre Pettis, continues to tout the fact that their users are their R&D arm. And in fact, this is true for one of the types of innovation listed above. The OSHW licensing of their product allows any user to take design files for the printer itself (or the printer the company was based off of, the RepRap), modify or improve it and then submit it back to the community. Perhaps a user is a circuit wiz who wants to work on one of the stepper motor drivers. He can simply download the design files, add any desired changes, test them and then resubmit them to the community. The OSHW MakerBot then can also roll these ideas into future products (as could competitors). In this way, the user who decided to work on the stepper motor driver made a subsystem improvement, an innovation that the entire community shared once it was complete.
The Open Source Ecology (OSE) movement is another one of my favorite examples. They have a top-down approach where a centralized team does all the dreaming on how to manufacture accessible, repaiarable and–most importantly–affordable industrial equipment. Then decentralized volunteers work on subsystems The idea being that these proven plans could then be taken and act as a “Civilization Starter Kit”. However, in all of this, we see that the OSE is actually a process improvement. Opening the plans and the manufacturing techniques to allow for lower cost and higher reliability equipment. They are not, however, planning or dreaming any piece of equipment that does not exist in the marketplace today. Rightly so, that is not their mission, nor the reason I highlight them. They are an example of a well managed centralized organization that is driving innovation, and doing so in an open manner.
So perhaps my last ditch hope at finding the requisite system innovation is one final “off the wall” organization that works in a similar manner to the OSE. DARPA (and sister organizations such as ARPA-E, etc) gives grants for ideas that are…well, crazy. They have provided funding for some of the most illogical projects and yet they push the limits of technology. They are centrally dictated, with the central body determining what systems are needed and who will receive funding (based upon proposals). They then award contracts for some or all of a particular system they are piecing together, with hopes that the final product will be a functioning systm. While they likely will not be proposing the next transistor (and its associated required systems), they are working on the technology of tomorrow that may end up trickling down to other companies once the ideas are disseminated or the patents expire. (note: this idea was posited in the comments section of my last article by steveo…thanks!)
So while I still lament the loss of a Bell Labs type atmosphere, it is simply a sign of the times. Large, centralized research work can happen, but in different sectors and situations than days gone by. Universities and software companies now provide the centralized research for long term invention and innovation. Distributed research helps to provide improvements in subsystems, processes and systems, depending on who is doing the research.
Have you seen other examples of distributed research? Do you believe we’ll drift back towards a centralized model? Do you work in either? Let us know in the comments!
Thanks to somethingstartedcrazy for the picture of the reverse waterfall…trickling up, if you will…