CloudFab is a new online marketplace that provides engineers, designers, and hobbyists access to a network of job shops who provide fabrication and prototyping services (currently 3D printing).

We mentioned the coming launch of the CloudFab ecology before.

Below, co-founder Nick Pinkston explains what gaps the project is aiming to fill, and Eric Hunting gives an analysis of the field’s present shortcomings.

Nick Pinkston:

“I’m currently only looking at the singular goal of connecting fabrication with end-users. To me, there are two main barriers of entry we must cross to allow people to directly create things: the design gap and the production gap.

+ For the design gap, I currently see three users types – each crossing the gap a different way:

– General Audience (e.g. my mother):

They want to have control to design / tweak things, but have no interest in learning any form of CAD software.

I see creator apps (a la Shapeways, NikeiD, etc.) as the way of crossing this gap from their ideas -> 3D file / specs. If they want to make a custom glass, they’ll find a creator that let’s them customize it, but they won’t design it from scratch.

– Casual User:

These are hobbyists who are willing to use simple design tools to design objects from scratch, but they won’t won’t accept a steep learning curve.

I see further variations on the themes of SketchUp, OpenSCAD, Aviary’s coming tools, etc. as being a good fit.

– Serious User:

They already know how to use CAD or are willing to do so with a slight nudge. They are the engineers, designers, and serious makers. They’re relativity well served by current CAD packages, except it’s expensive pricing and oligarchic tech sharing (from plug-ins, algorithms, sub-assembly part libraries, etc.).

+ For the production gap, I see a few issues:

– File Processing:

We need to make sure the files are ready for production. There are a lot of variables that inhibit certain designs such wall thickness, overhangs, internal geometries, etc. This isn’t a problem that been solved well so far.

– Quality Control

With so many printers out there with many variables (engineering, environmental, etc.) that currently have no feedback control in the machines – it’s really more art than science in most cases.

– Functionality

Making parts that stand the rigors of daily use has been difficult for most processes. Some, like SLS, work pretty well though. There are also issues of rough finished parts working well together without manual intervention. Finishing could be its own category really – most processes require some type of intensive finishing in practical usage or to look good.

– Assembly / Logistics

Most products aren’t one material, and most use things that aren’t currently fab-able like LEDs or even nuts & bolts. Some how the supply chain has to pull these together and either assemble the end product or go the IKEA route and give simple assembly instructions.”

From Eric Hunting’s response:

“This is a great but ambitious project and has some goals in common with some other projects going on lately, such as 100k Garages;

http://www.100kgarages.com/

I get the impression that, in general, there seems to be emerging in the open manufacturing/Maker community this sense that Metcalf’s Law isn’t quite kicking-in yet given the ad hoc way people are communicating at present and that a more organized network and centralized repository for information and project management is needed to catalyze a real open manufacturing revolution. I’ve touched on this idea a few times myself. When I proposed the Open Source Everything project in TMP2 (The Millennial Project 2.0) several years ago I discussed the need for a next-generation SourceForge and, again, that concept has emerged in my discussions on the ToolBook Maker/Author-cooperative concept -which seeks to concatenate all civilization’s industrial knowledge in open source forms while also creating a simple engine of economic support for those doing this work- and its related notion of the Vajra Maker-incubator eco-village. But it looks also like we may be looking at a near-terrm War Of The Search Engines here as competing approaches to this roughly similar idea go through a phase of Darwinian competition rather than cooperation -which I suspect relates to the simple fact that none have quite reached critical mass and thus a broader community consensus. (as an elderly venture capital broker once told me, ideas are like ox carts. It’s doesn’t really matter how nice or well made the ox cart is if it’s standing still. People will hop-on to the ones that are moving and look like they’re going fastest in the direction they want to go)

I think Nick has the right idea in his definition of the three basic tiers of design ability/interest in the society at large. At the level of the Serious User price is the key issue and I think the key factor in that is developer’s lack of understanding of their own products’ market potential -largely because their perception of their own markets is based largely on assumptions. Many companies that produce technical/industrial oriented products and have long been selling in what is essentially a ‘business to business’ market price things very high on the assumption that their market is composed of a very small number of possible customers who amortize their expenses over anticipated mass production. (or professional/technical service for companies who amortize their expenses over mass production) In other words, they assume their market is very small in number of customers, very high in value per customer, and they price according to this assumption to cover their development costs -never realizing that they could be locking-out their actual potential market because of that price. The idea that there could exist a broader market -larger in number, lower in value- of people who use their products to make things for themselves on-demand is often utterly beyond their imagination. So the solution is open alternatives and/or proving the existence of the larger market. (and somehow getting that information through the plastic bubbles and the thick skulls…)

For the General Audience and the Casual User the traditional solution has been modularity; compartmentalization and encoding of complex knowledge and information into simpler but still flexible reconfigurable elements of standardized form and interface. One of the curious phenomenon in the evolution of the PC has been that the modularization and standardization of components actually resulted in a great diversification of the ultimate form of the PC -because modularization with standardization of interface enables customization with a reduction in skill/knowledge. You can devise very clever and novel designs for computer cases without having to know how microprocessors work or how to make a motherboard as long as the basic standard functional parameters for a computer case are clearly defined and the form factors of standard components offer sufficient topological flexibility.

Today, however, software has offered us a new approach to this simplification of design process through encoding of information in the form of procedural design software. In other words, design based on procedural modeling systems as demonstrated by such things as the creature creator tools of the game Spore. In the realm of mass customization theory such systems are called product configurators and perform the same function of modularization through a standardization and hierarchical structuring of design parameters rather than a physical compartmentalization of component elements and interfaces. In practice, though, the technology remains crude and limited to discrete products -nothing in the manufacturing realm is even close in sophistication to the Spore creature editor- because we haven’t really gotten too far -or deep- yet at a comprehensive ‘proceduralization’ theory or model for most manufacturing processes. (though there are folks in open manufacturing working on this, as with the OSCOMAK project http://oscomak.net/) Near-term, though, modularization and the use of modular building systems is likely to remain a key simplification approach and will rely on our Serious User tier’s ability to invent good modular systems and cooperate/collaborate on their interface standardization, just like the commercial parts developers for PCs.

On the production side of things, I think a key and overlooked issue in the logistics of currently rather dispersed independent production is shipping overhead. We have people jumping on the bandwagon, buying their own fab tools, setting up shop on-line (hoping to cover the cost for their personal use by making the hardware pay for itself), and then wondering where the customers are because they forgot that location still matters and shipping costs limit casual use. At the tier of the General and Casual user, the accessibility of custom production for modest items is limited by the cost of shipping, which hampers experimentation. This represents a critical gap between the on-demand fabricator and the designer/end-user and there don’t seem to be good solutions. The cost of using a service like Ponoko, for instance, doesn’t make a lot of sense for a great many users and items. You need to be able to order many units of small items -more than you might need for yourself or for an experiment- or a complete made-to-order product (or kit) to make the shipping cost viable or, for a very large bulky item, need to have a single article of very high value. This is a key issue for independent production that I foresee will produce a trend where commoditized materials and modular items of higher unit value will be more common in production for distribution over long distances and the more custom production will become increasingly local. (trade goods vrs. personal goods, as in the pre-industrial period where the design and intrinsic value of goods was keyed to the distances they had to be transported and the methods of transportation).”