Having to build an apparatus, trying to show the concept

Question

I have to build an electro-mechanical device that has a few sliding and rotating parts controlled by servos and actuators from Phidgets.

I made the prototype with wood; it was crude, imprecise, and looked like it was built from scrap from a junkyard, but it proved the concept.

Now, we are going to need to make the real thing. I spoke with a local company that does CNC work and I tried to explain to them the concept. But I suck at drawing, and the original prototype covers just a part of what we need.

I have several options that I am contemplating:

• Building something out of wood again to show them what I need. But it was time consuming and not very good last time
• Build something, without the motors, out of legos to demonstrate the moving parts.
• Make a proper model in SketchUp, but it won't be able to demonstrate the moving parts. The CNC company told me they'd be happy to see SketchUp/Autocad files, but I'm afraid this won't convey how things move during the operation.
• Writing on Stack Exchange to get some fresh ideas :)

What would be the best way to describe a device to people foreign with the concept? They need to understand how it operates, the constraints for the moving parts, etc and eventually come up with the final schematics. So I need to get the point across.

I have never done this before, and this is domain that is foreign to me.

Answer 1

I think that it depends upon what your purposes are.

If you're looking to pitch the concept to investors, then there are a couple of routes to consider.

The most obvious route, as you mentioned, would be to build a second, more refined prototype. It may be worth giving this another consideration despite the investment of your time. Nothing sells a concept to an investor quite as well as an actual, functioning prototype.

For perspective, Dyson claims that he went through over 5,000 prototypes in designing his first bagless vacuum.

Another route to consider is that sketchup appears to support creating animations. If you eventually want to ship this concept off to a machine shop, then you'll need to invest the time to create digitized plans. This is likely the fastest path to having a refined presentation of the concept.

One other concept to consider is to create a video recording of the existing prototype. That will both demonstrate the moving portions of the concept as well as digitizing the concept so it is more easily shared.

Rebuilding the prototype in a non-functional way (leaving out motors) with a material that you never intend to use for production doesn't seem like a great approach. That seems like a step backwards from where you're currently at.

Another option to consider is to contract with a mechanical design shop. Not all CNC shops have mechanical design capabilities, so you may need to shift who you are talking with at this point. A design shop should have an engineer or draftsman that can work with you on the concept and then build a specification. It's worth noting that this can appear to be an expensive option, but you are paying for their expertise in creating schematics.

Answer 2

The best route in this situation is a simple feedback loop:

The fact is, you don't need to have amazing drafting skills to have good ideas, nor do you need to have amazing machining skills to have good ideas. You do need persistence.

Take a video of your device (if you still have it), and have someone make the CAD file. It will come out wrong. Redo, over and over, until it comes out right. Why CAD? It's cheaper to make CAD files then it is to make actual prototypes. Then make prototypes, over and over, until it comes out right. Then make production parts. Test them over and over, with QC checks, until you find better ways to make them come out right without defects.

Looking at it like "I need a definitive solution for it to come out right the first time" is how scientists approach things. That's what makes engineers different. Here, we modify and redo over and over until it comes out correct, not just settle for the "one answer".