Disclaimer: I'm a practicing mechancial design engineer and opensource hardware contributor (http://blog.OpenBeamUSA.com).

A commercial CAD software package is really 3 components:

1) A component to create your parts (Part design)
2) A component to create assembly from your parts, and move / mate them for interference detection
3) A component for exporting the previous 2 parts into a paper drawing / engineering file exchange formats

Of all the OSS options, FreeCAD is probably the closest one to being able to implement all these. Assembly module is still under development, but from what I've seen of the part modelling module, it's done really well and it's the closest to commercially available close sourced alternatives. (I own a seat of Solidworks. At $5k per seat, it is not for hobbyists).

If you are a student, Solidworks, AutoDesk Inventor, ProE (now PTC Creo), all offer great discount student packages. Be warned though - some of them are only 2 year licenses.

For the lower budget end, there's Alibre Design as well. No experience with this.

The 3D printer community also uses OpenSCAD a lot. Personally, that's like using a butter knife for brain surgery incisions. OpenSCAD uses a scripting language to script a part into existence. I've heard that there are now graphical front ends to do this, but have no experience and no desire to learn it.

http://reprap.org/wiki/FreeCAD
http://reprap.org/wiki/OpenSCAD
http://free-cad.sourceforge.net/

I'm a small business owner (I created OpenBeam: http://www.kickstarter.com/projects/ttstam/openbeam-an-open-source-miniature-construction-sys). It is basically a small, nice version of an erector set, that is currently being used for building 3D printers. (See: http://reprap.org/wiki/Kossel).

US manufacturing is *hard*, for sure, for small businesses. In fact, the system is set up so that I'm better off shipping jobs overseas.

We buy our extrusions from a small mill in California, a family owned business. Our first batch was great. We made a small engineering change on the next batch and ordered the extrusions in October of 2012. We received the parts in early December, and the black anodizing was crap - it literally looks like it's been dive bombed by seagulls with diarrhea. We shipped back 700 of 2000 pieces for rework, and we still have not received it back. Meanwhile, I'm out of stock, I have thousands of dollars of backorders that I can't fill, and I still have no idea when I'll get replacement stock back in. And to make things worse, when we complained initially about the quality of the parts, the answer we got was literally "you're small potatoes, we don't have time for you"

Meanwhile half way across the globe, my injection molder (http://blog.openbeamusa.com/2012/05/18/behind-the-scenes-injection-molding/) is churning out parts, 50,000 at a time. He always delivers when he says he'll deliver. With UPS and Expeditors I can get goods landed on my doorstep anywhere from 48 hours to less than 3 weeks for ocean freight shipping. It costed me $1000 to ocean freight half a metric *ton* of parts, and it'll be here in 3 weeks. The reason for going overseas for injection molding is simple: The material we use is a high end glass-reinforced nylon and the only shops the US that can handle it are military and aerospace molders and they demand an incredible premium.

On top of all this, I currently import a bunch of motors, pulleys, bearings for my 3D printer kits, US customs requires that I file an individual HTS classification for each line item, and taxes me individually. I then pay my old coworker's kid $20/hr, which is a princely sum for a 14 year old girl, to do my packaging and kitting. However, If I paid some guys overseas $10.00 a day to do the same job, I can declare my imported goods as "construction toy set" and avoid paying import taxes all-together. Therefore, there are absolutely NO incentives for me to keep the packaging job in the US, except for the short flexibility between an engineering change and getting the change pushed through on the line.

When it comes to export, I'm equally screwed. Until I signed up with Expeditors, there was no easy way for me to export my shipment around the world. So while I have customers in the UK, EU, and NZ/AU areas, for the longest time I had to resort to USPS Priority mail to ship them stuff, and priority mail rates just went up. Surface parcel service was discontinued a few years ago during budget cuts, so unless you are a bonded importer / exporter, you really have no option of doing a low cost export. Meanwhile, I paid US$20.00 for a batch of parts for 2 day shipping for a crate of timing belt pulleys from Shanghai to Hong Kong. There are so many Chinese logistics company these days that shipping is incredibly cheap to move things around in China.

People don't realize that the world is getting a lot smaller these days. The other day a vendor returned an email quotation - 5 weeks after initial RFQ. I had already paid someone else and landed parts in that amount of time. A supply chain is only as strong as its weakest link and it seems like for small businesses there are just no good options for manufacturing.

-=- Terence

Wrong. FDM is used commercially to produce parts - including airplane parts. (I know, because one of my (OpenBeamUSA.com) vendors is printing cosmetic trim pieces for the Airbus A380). Commercial systems have no problem with overhang because they use a water soluble support system. And if you want something prototyped out of production representative material, it is one of the faster (and hands off) way to do it. I've used FDM for Ultem, Polycarbonate and ABS parts before.

The reason you don't see this in the hobby market, is because Dimension's patent on water soluble support, and heated build chamber, is valid until 2014 (or 2016). That's why there is no hobby grade 3D Printer on the market that is completely enclosed and they only have a heated bed. Most people enclose their printers if they are printing with ABS for a variety of reasons.

-=- Terence

Posted also on the Kickstarter comment section:

Kickstarter project creator here: I'm the guy behind OpenBeam (http://www.kickstarter.com/projects/ttstam/openbeam-an-open-source-miniature-construction-sys).
And in case anyone's wondering - we shipped the majority of our rewards a *month* before the original promised date. That probably puts me in the top 5 percentile of projects...

Let's take a look at the new rules one by one:

“What are the risks and challenges this project faces, and what qualifies you to overcome them?”

- Okay, this is perfectly valid. I am surprised KS haven't done this earlier, because there are quite a few clueless guys (http://www.kickstarter.com/projects/277210494/paint-be-gone) - *(http://www.kickstarter.com/projects/277210494/call-key) and http://www.kickstarter.com/projects/277210494/key-pad-case) out on here who seems be doing the "throw s*** at the wall and see what sticks" model of development.

"Product simulations are prohibited. Projects cannot simulate events to demonstrate what a product might do in the future. Products can only be shown performing actions that they’re able to perform in their current state of development."

"Product renderings are prohibited. Product images must be photos of the prototype as it currently exists.
Products should be presented as they are. Over-promising leads to higher expectations for backers. The best rule of thumb: under-promise and over-deliver."
Okay, so KS want a working prototype. I get that; that's pretty straightforward. But it doesn't stop someone from *faking* a prototype on camera. This however, won't stop a project like iCase (http://www.kickstarter.com/projects/1791911961/i-case-iphone-4s-and-iphone-4-bumper-case) from being a train wreck, as the protoytype would likely have been SLA, painted, and the problem wouldn't have been apparent until the metal parts were CNC machined and fitted and found to short out the iPhone's anntenna.

The OpenBeam project would have passed these requirements; we had a physical prototype for shooting the video, as part of a good product development practice.

"Offering multiple quantities of a reward is prohibited. Hardware and Product Design projects can only offer rewards in single quantities or a sensible set (some items only make sense as a pair or as a kit of several items, for instance). The development of new products can be especially complex for creators and offering multiple quantities feels premature, and can imply that products are shrink-wrapped and ready to ship."

And how would KS define "Multiple copies" of a reward? This I have a problem with. When you're in production, you are trying to get the manufacturing volume up to bring the costs down. If I were launching OpenBeam now, would I be limited to selling one stick of aluminum and one of each bracket to my backers (who wouldn't be able to do anything useful then with this?) If I packaged it up as a "kit", like I had on my KS, would I have gotten around these restrictions? Who decides whether multiple copies of the same item is required for the item to work (ie, construction toy kit), and when it becomes a way to side step your rules? How much "individual judgement" is there to allow the listing of a project, and do you consider the project creator's background (ie, having successfully delivered on a previous project) when you allow them to post? With the amount of controversy about what gets allowed (*cough* Tangibot (http://www.kickstarter.com/projects/mattstrong/the-tangibot-3d-printer-the-affordable-makerbot-re)*cough*) and what doesn't on Kickstarter already, this rule is probably going to make your selection process more Apple App-store like (arbitrary with no recourse for the project creator if you are not selected).

(Edited to add: The real problem, that KS probably don't want to admit, is that none of their hipster workers have a sufficient engineering / science / technology background to effectively monitor and approve Kickstarter projects to begin with. That's why they let stupid shit like Z-Torque (http://www.kickstarter.com/projects/1202837765/z-torque-bicycle-cranks-ride-faster-longer-easier?ref=live) in, but plenty of GOOD projects get rejected).

-=- Terence

There's two issues here: There's the issue of whether current (or more stringent) security measures can still be beaten by a determined foe, then there's the issue of actual Li-Ion batteries going kaboom. I'll address the later.

Li-Ion batteries are some of the highest energy-density storage devices available to the general public. As such, they *are* dangerous. I design battery packs for a living, and let me tell you - if not for microprocessors and safety circuits, we wouldn't use Li-Ion batteries.

They are the only batteries that I know of that can fail dangerously when over-discharged. You start creating internal shorts of lithium whiskers between the cathode and anode, which bypasses any cell safety circuits.
They go boom very spectacularly if you overcharge them. The model RC heli folks have found this out the hard way, as they tend to run bare cells without protection circuits to save weight. (http://www.youtube.com/watch?v=hcwOwf55Rtc).
They have very low internal resistance, which means in a short circuit, they can release energy very quickly. Every manufacturing engineer at the company I work at have welded calipers to cell tabs, from accidentally touching the wrong stuff while taking measurements.

For a good cell manufacturer - and I'm talking about the LiShens, Sanyos, Kokams, and Panasonics of the world, the failure rate is 1 in 1 million. It's just a fact of life. The fly-by-night operations in China, responsible for some of the god-awful counterfeit cells out there, god knows what those failure rates are. And the vendors who use these cells tends to not put in the safety features (look up a BQ20Z70 chip, for example) to make a failure more likely.

The nightmare scenario would be some dude getting some last minute work in at the terminal, plugging the battery in for charging. Then the plane takes off with the laptop in the overhead compartment where the oxygen lines for the safety masks are kept, and the cells let go. Judging from how much energy a single 18650 cell can contain, it could easily do some very serious damage.

With the prices on Li-Ion dropping and more devices using them, it's no wonder that almost all of the 22 incidents reported occured in the last 3 years. Still a small number considering the amount of airplanes in the air at any given time, but enough for someone to pause and think...