MIT's Eric von Hippel on open innovation models

Famed MIT open innovation researcher Eric von Hippel has a new paper out on open innovation models, of which the DIY Drones dev teams are one. It's a pretty academic read, but I thought the framework he identifies in the above diagram was very interesting. I've drawn what I think the path of autopilots has been over the past five years ago, as the ease of communications between distributed participants and the falling cost of autopilot technology has taken what was once a military-industrial technology and made it accessible to open source teams such as our own. Did I get the arrow right?

Views: 437

Comment by automatik on December 14, 2009 at 11:16pm
What's the title of the paper?
Comment by bGatti on December 14, 2009 at 11:30pm
This curve shows Design costs coming down at the same rate as communication costs. In response to the question about the curve being right; I would pin the curve steeper on design costs, and almost unchanged on communication in the last 5 years - since the means and cost of communication are largely unchanged since 2003. Circuit geeks have been openly sharing designs using bulletin boards since 1980's or so? The increase in participation (which might be confused as lower communication costs) are I think more related to the design costs side lowering the bar for playing with mcu generally and RC flight in particular.

The graph reinforces the view that Open Source is increasingly relevant as design becomes a significant component of realized cost - nearly 100% in the case of software - somewhat less for electronic boards (ardu) - and least for brute hardware projects like say Open Source Ecology
Comment by Garry Qualls on December 15, 2009 at 12:06am
I believe he is referring to paper #16 on that page. "Modeling a Paradigm Shift: From Producer Innovation to User and Open Collaborative Innovation"
Comment by bGatti on December 15, 2009 at 12:48am
On wading through the paper; I'm struck with the existential conflict between falling economies which employ patents (the west), and rising economies which ignore them (china). Add to this the role of communication costs in shaping which projects are viable as open or closed, and it occurs to me, that the ideal mix of open and closed would provide enough collaboration to complete the project, and yet enough propriety to command a premium in the market, and that to some extent a language barrier may provide both.
Comment by Garry Qualls on December 15, 2009 at 1:09am
I would suggest that communication costs have continued to come down as social networking software has appeared, improved, and been adopted by mainstream internet users. I think new people feel more comfortable clicking around on this style of site site than they do wading into the boards at rcgroups.

As to arrow direction, the commercial autopilot I am most familiar with started with a single-user innovator at a university who went to work for a Producer Innovator. The Producer Innovator made the decision to open-source the autopilot software so that they could sell the hardware internationally, so that particular autopilot went down and to the right. I realize that does not directly relate to Chris's question.

I guess I might disagree with putting these two costs on orthogonal axes, maybe? I think that the cost of a single good electrical engineer designing something as complete and nice as an Arduino Duemilanove from scratch has not changed all that much over the years. Fab and component prices are on a well-known downward slope. The extra reduction of design costs has come from sharing designs in the Open Collaborative space so that no one has to start from scratch and innovations are not limited to electrical engineers. The availability of common file formats, hardware design tools and software interfaces fit in there somewhere too. So I guess I am saying that the open collaborative tools that are being used by the Open Collaborative Innovators reduce communication and design costs in almost equal measure. The collaborative communities also tend to assimilate the Single User Innovators and Producer Innovators that aren't protecting intellectual property or complying with ITAR. So I guess I vote for a red arrow that goes a little to the left but mostly straight down...
Comment by bGatti on December 15, 2009 at 2:44am
@Gary,
yours is an interesting point I think not covered in the article, which suggests that communication can itself lower design cost by reducing the duplication of work; which if accepted makes these costs interdependent rather than orthagonal (your next point).

As for this framework, I think the design costs have come down as such:
1. GPS was unlocked and consequentially commercialized and miniaturized. Lighter and cheaper.
2. Gyros have become MEMs, thus lighter and cheaper than spinning gyros by like 3-4 orders of magnitude
3. Lithium and ESC have improved the lifting capacity of small airframes - while the components have become lighter, cheaper, and more accurate.

So I see these component costs having reduced the cost of design specifically (and production), meanwhile I don't see a cost difference between one BB or another in terms of collaboration (in the last 5 years). I suspect most collaboration is via usenet groups, email, and their modern equivalents.

It just seems obvious to me that if you were to roll back the technologies I listed to their previous state, Autopilots would be massive, and only relevant to fullscale airplanes - but if you rolled back say this blog site to its Usenet progenitor, the cost effect is demonstrably minor as thousands were collaborating on Usenet.

That said, the article is about trends, and their future effect on diverse corners of the economy; Autopilots do not fit into the graph well because their primary constraint is not design or communication, but regulation, and there is no axis for that.

I am not convinced that we have common tools: there are relevant mcu innovations from Cyprus, Freescale, Atmel, ST, Arm and Microchip (ST mcu has ESC driver as peripheral, Cyprus include PGA which could read thermopiles directly, Microchip has mcu's with matrix multipliers and radio capabilities. Etc...) and yet, each project must choose one mcu brand and exclude the rest - good engineers probably should be up to date on several mcu brands - these are a long way from common "file" formats or tools from my perspective.
Comment by automatik on December 15, 2009 at 10:08am
There is a nice list of papers pointed by Chris's link, however I think that post refers to
"Modeling a Paradigm Shift:From Producer Innovation to User and Open Collaborative Innovation"
direct link to PDF paper:
http://papers.ssrn.com/sol3/Delivery.cfm/SSRN_ID1513273_code698198....

hopefully this helps others finding the right paper :)

3D Robotics
Comment by Chris Anderson on December 15, 2009 at 10:11am
Good catch. Apols for the bad link before--post now corrected
Comment by Patrick Egan on December 17, 2009 at 10:33am
I have to agree that it’s definitely not just the autopilots (or the software) that have an effect on the cost. Many of the innovations in the “electric revolution” have brought the costs, complexities and proximity issues of RC/RPA down a few notches. Is a “mishap” from corrupt or malformed code a factor? How can anyone guaranty that the system will work? Are we factoring in all of the hardware purchased and time spent on past and present iterations/versions that didn’t/don’t work? Then there is the regulatory side i.e. certification (whatever it looks like) is going to add thousands to the cost of using/producing any autonomous capabilities. The manned stakeholders really don’t see “open source” as a viable candidate when it comes to certification for use in the NAS. So, who will want to pay $10,000 for an autopilot (demand) that can only be used in a very small envelope?

Comment

You need to be a member of DIY Drones to add comments!

Join DIY Drones

Groups

Season Two of the Trust Time Trial (T3) Contest 
A list of all T3 contests is here. The current round, the Vertical Horizontal one, is here

© 2019   Created by Chris Anderson.   Powered by

Badges  |  Report an Issue  |  Terms of Service