Wiki Design: from Toasters to Spaceships
Participatory Exploration. Frednet. Lunar Boom Town. Open Luna.
These all deal with the concept that we are trying to take the brainpower of the interested public and use it to solve the technical, political, and business problems that confront our efforts to expand into space. Consider a tool that can facilitate this.
Look at Wikipedia. According to last year’s annual report (check it out in your spare time, Nick,) there were “approximately 100,000 active editors (defined as users who made more than 5 changes in the last month).” 100,000! That’s a huge number of people!
With the 11 million articles on Wikipedia, you can be sure that many of these editors are fueled to participate in a wide range of articles by the synergistic combination of articles that they can work on. In other words, editing in Wikipedia gets “sticky” (Check the definitions at the bottom)
So here is the point. Crowd Sourcing is good. Better Crowd Sourcing is better. Better Crowd Sourcing can be had by implementing a dedicated web based methodical structure that fosters and requires attention to the essential questions of systems design.
So I am hatching this idea for something that could be a Wikimedia project, specifically for designing things. It would work a bit like this:
You come to the wiki design sight and tell it that you want to start a new design. It asks you some basic questions like what your primary objective of need is, what kind of system it is (Vehicle, building, processing machine, etc.), Does it require data processing, etc.
The site shepherds your thoughts into a rudimentary top level systems architecture framework by asking you questions like: what does it do? And how might it do that?
It gives you some templates for functional and physical breakdowns, templates with high level headings for a system specification document, and you, the user get as detailed or a vague as you want at this point.
So then your site is live and anyone can come in and populate the content, like with Wikipedia, but unlike Wikipedia, some powerful organizing tools and templates are integrated with the content.
Some of the possible features:
- Integrated 3D modeling web app that helps with part numbers and hierarchy of parts
- Expired patent and journal search that lets you link relevant patents to functions or subfunctions
- Discussion and voting tied to specific elements of the system definition.
- Commenting on parts of the system definition (Saying things like: “This design is horrible. If it were 3 inches long it would have way more strength and only add a small amount of length”)
- Chat with other members of the project
- robust and targeted permissions to set “baseline” requirements, functions, components, interfaces, etc
- Automated quality check that alert users to possible functional overlaps, shortfalls, etc.
- Autocheck to make sure that users don’t give functions titles that are nouns or verbs as titles to items.
- Freedom of Information Act Request facilitation.
- Reuse of components, functions, etc from other projects. (Got an idea for something with wheels? Pick from a myriad of projects in which the wheel was defined already!)
The idea is that most people don’t know beans about systems engineering, requirements, or interfaces. Design by committee, forum posting, voting, or by blind feel with no knowledge or application of systems engineering is not an effective method of harnessing the domain knowledge that many people do have.
So who is with me? Let’s storm the Wikimedia foundation and get them to put this thing online so that we can go about the business of designing space vehicles in style!
9 Responses to “Wiki Design: from Toasters to Spaceships”
Leave a Reply







Robbie Schingler on June 21st, 2009
Hi John,
This is a great idea! As with any website / on-line tool, it is generally useful to start with a very simple site. I would suggest starting with something that people already do ineffectively, but then provide a tool to make it easier for people to get their job done. As this catches on, you let the community innovate within the bounds you give them, then slowly start changing the boundaries so that you have a full-on project incubation / open source hardware tool.
From a NASA perspective, I believe that we share our ideas and concepts in a linear, political way. I believe we can more effectively incubate projects if we do it openly in a trusted, crowdsourced way. This is the motivation and basis behind the Side Project App I blogged earlier.
But what are your ideas for getting Mediawiki onboard?
–Robbie
John Benac on June 22nd, 2009
One way to go about it would be to look at some of the myriad of systems engineering applications that are already in existence, such as CORE or DOORS, and put them online, with the added features only offered by the internet (chat, user input function and component libraries, etc)
http://www.paper-review.com/tools/sas/read.php
1) Crouwdsource the software itself and then approach the Wikimedia people in San Francisco and propose integration with their larger suit of tools.
2) Or, alternatively, build a loose coalition supporting the idea on the net and work to define the software (PowerPoint style) to such a detailed extent that we can simply go to the Wikimedia foundation and have them accept the idea (and perhaps a few members of the coalition as employees) and build it themselves. They spent 1.3 million last year on “technology”
3) A third route (and the one that would yield the poorest quality, it think) is to download the MediaWiki source code software and adapt it for a more robust application.
http://www.mediawiki.org/wiki/MediaWiki
4) Have Wikimedia lead development on a crowd sourced software development.
Rolando Quintanilla on June 25th, 2009
John,
I really like the idea. Is MediaWiki the only option for going to with this idea? I mean other companies such as the famed Google might be able to profit considerably from this type of venture. I think that even Amazon could open up markets from this. I see people designing things through crowd sourcing. Google could use it as a means of advertising to the users, and as a service to companies. Amazon could sell what is made on their website.
Actually, now that I think about this more, these companies that emphasize marketing metrics can work on technologies that help you meet your specification with the “right” components. The amount of development that goes into fitting people with what they would be most interested in purchasing is not simple, and the work that has already beend one can be applied to helping companies/organizations meet their specifications automatically (or at least lower the amount of research time).
-Rolando
John Benac on June 25th, 2009
I see Google as a company that produces tools, and MediaWiki as a entity that is a data library.
When you consider the web traffic that Wikipedia commands on the internet (and its rankings in Google searches,) I believe that the two companies are have a symbiotic relationship.
So Google might be a company best apt to develop tool (integrated Google Sketch Up, target adds and content from suppliers relevant to the design, etc,) but I don’t see them (Google) hosting the tremendous amount of data that would be generated by the “croudsourcers.” Look through their labs. They will store your content (sometimes with a fee, like in Picasa web albums,) but YouTube, which they acquired, is the only place where there is a significant amount of free storage that is given high bandwidth to distribute to the public at large. Remember that Google Video, their own solution to online video sharing before they aquired YouTube, did not allow you to upload content for the masses.
Wikimedia is all about getting your data and giving it to the world, but their tools are not so detailed as to enable the kind of robust tool the is necessary for the type of serious design tool that croudsourcers would need to go beyond a glorified bulletin board.
So in the end, Google and Wikipedia have demonstrated their strategies, and neither company is wired to take this concept and develop it (except for something coming out of Googles fames “20% time,” but you cant bank on someone doing this as a side project.)
I think that the way to go is to go open source with the software, perfect it, maybe make a tiny little company with it, and go get acquired by Google or Wikimedia.
How many countless examples of successful open sourced software projects have we seen? Open Office? Wikipedia itself? Mozilla? Linux?
So, the abridged version, 1) define the software itself, 2) have it developed “open source,” 3) Sell it to Google or WikiMedia to support the bandwidth and data needs 4) Change the world
John Benac on June 25th, 2009
So I’m taking this ball and running with it.
Open Source? Done.
http://wikidesign.sourceforge.net
Sample Output of the process? Done.
http://4.bp.blogspot.com/_bOx9.....ecture.JPG
Dedicated Blog? Done.
http://design-wiki.blogspot.com/
We’ll see where it goes. Thanks for the encouragement.
Rob Freeland on July 10th, 2009
A few comments.
Collaborating over the internet to write what’s essentially an encyclopedia is one thing, since it’s all words and data, and released on the same medium used to create it. Working on a spaceship is orders of magnitude more complex. First of all, you don’t have to worry about loss of crew if you misspell their last names.
You are going to need to do physical prototyping and testing to validate the correctness of your models. Who’s going to volunteer a metal shop and materials to prototype? Who’s going to volunteer fuel? A launchpad?
And people with a knowledge of systems engineering don’t necessarily have the domain knowledge necessary to correctly define a system architecture. Look at how Ares I has changed from its original form in ESAS to now. Assumptions made during ESAS were found to be incorrect, which forces re-design and re-evaluation of the architecture. How do you know you’re not making the same mistakes?
I would focus more on figuring out how to capture the rules of thumbs of the domain experts at the systems level. How do you capture what a welder knows about welding, and how can that influence a design at the system level?
Also, just like Wikipedia, but at a larger scale, you’re going to run into a situation where two experts firmly believe their design for a component is the best. How does the audience decide? If I’m a thermal guy, should I be able to influence the design of the structure?
For this project, how are you going to get the investment of people from all of the design domains necessary to design a spacecraft? Wikipedia may get one committed grad student to fill out their page on quantitative thermodynamics, but I bet there are thousands more that troll Joss Whedon’s entry, looking for mistakes. How are you going to cast a large enough net to make sure your areas of design are covered?
I’m not knocking this idea. I’m not convinced it wouldn’t work. I just wanted to ask some leading questions to push this idea past its “Gee-Whiz” stage.
Rolando Quintanilla on July 11th, 2009
Rob,
Your comments are very good. I think that you are stating that it would not be feasible to develop and then use the architecture John is proposing for the design of complex systems like spacecraft, because no one will be willing to work for free to get the job done. Now, I don’t think John is suggesting that we develop a spacecraft from the collaborative effort of volunteers. What I think John wants to do is spear-head the development of software that could be used by individuals, organizations and companies to complete projects more effectively than what we currently do. These organizations and companies will have to pay to get their products developed.
I do question the feasibility of having Wikimedia store all the data. I don’t think they have the resources to do this, and their model of business implies that they won’t have the resources for a very long time (hard drive space, bandwidth, staff costs a lot of money). From recent discoveries I do believe that MediaWiki software can be used as a portal for bringing all the data together. Recently, I found out about MediaWiki Widgets (http://www.mediawikiwidgets.org/Main_Page), which allows users to embed things like YouTube videos, Google Docs, SlideShare into the wiki. The code required to display the content is still very simple and appears to not require significant resources from the server running the MediaWiki. Now if we could get more Widgets developed for MediaWiki that are taylored to John’s idea, then we have the beginning of something.
A really good Wiki that already is using Widgets is OpenWetWare.org (http://openwetware.org/wiki/Main_Page). This group is a group that I think NASA groups should be looking at joining. Their mission is : “OpenWetWare is an effort to promote the sharing of information, know-how, and wisdom among researchers and groups who are working in biology & biological engineering.” . From first glance they appear to be building a very good infrastructure for collaboration online, with regards to the biological sciences. If you go to the website it appears to be very well layed out and planned. I am researching this group because it appears to be a very good model of how to collaborate online (or at least they will get there).
I hope to learn something about the way they are set up to help better organize Tx/Rx Labs (www.txrxlabs.org) web presence. Tx/Rx Labs is a hackerspace in Houston, which can be seen as a open community lab. Basically it is a group of members that pull resources together to be able to afford large hardware needed to complete projects (such as CNC machines, Laser Cutters,etc.) to build things that we couldn’t build otherwise on our own. The mission of this group goes hand-in-hand with John’s idea, and it even attempts to address your concern “You are going to need to do physical prototyping and testing to validate the correctness of your models.”, but for small stuff at this point. Our group is relatively new (a few months), so we still have to go a long way. Other Hackerspaces that are much further along and are really doing some cool stuff are NYC Resistor (http://www.nycresistor.com/) and Noisebridge (https://www.noisebridge.net/wiki/Noisebridge).
Anyways, I have digressed.
Rolando
John Benac on July 12th, 2009
So regarding feasibility… On a basic level,
Systems architecture is the new (~60 years) discipline of describing an architecture in a standard way.
Web technologies have recently progressed (~10 years) to enable an interconnected web based system architecture creation and dissemination tool
This tool does not exist currently.
A dozen client-only system architecture tools have existed for some time now, and posses the core desired functionality.
Regarding the feasibility of describing detailed system elements such as a design of an ablative tile that imparts an acceptable vibrational signature on the display screen in a cockpit; its all a matter of the quality of the inputs.
Configuration management is important in any complex system. So is the inclusion of decision gates and “freezing” the design in certain subsystems and at certain times. These are all issues that can be accounted for (possibly as optional options for more serious projects.)
While you wont get a multi-billion dollar design effort releasing drawings for the next Moon Base, you might see the uncorking of mature technologies that have been reinvented every time that a new “proprietary” program gets revved up with people who didn’t work on the last one.
A rising tide raises all ships, and WikiDesign can raise the tide of knowledge for everyone approaching the fundamental problems in designing systems bound by fixed constraints, such as spacecraft systems, transportation systems, medicine, etc.
Systems engineering starts with defining the system, then moves to building the system, and finally finishes with validating the system against the original design to ensure quality. WikiDesign would probably do well to focus on the first part, although as Rolando points out, croud sourcing the hardware is too far behind, although he did leave out one of my favorite groups: http://www.espacecenter.org/
pan on June 8th, 2010
I have been trying to do this exact thing for some time now. It takes more people than i seem to be able to get involved. Forget wikimedia, they are useless and hate “original research.”
old wiki
http://ni4d-issues.rbefoundati.....=Main_Page
new wiki
http://www.rbefoundation.com/w.....=Main_Page