Just IN- The Future of CAD/CAM Software

[kackermann]

I hope this is posted in the right place...

Here is a Web-hosted design tool that I created for designing and manufacturing Front Panels.



Efficient, functional, appealing tools that run inside every popular browser on all major computer platforms.

Soon, nobody will own design software, but everyone will have it.

Do you want your application delivered the same way? Call Keith Ackermann at (503) 829-6400, or e-mail kackermann at molalla.net

Note: The above link is a testbed. The Order... button is not functional.

[ger21]

Personally, I won't be using any web based software until it's the only option available.

[kackermann]

Personally, I won't be using any web based software until it's the only option available.

I find that very interesting. Could you elaborate?

I busted my chops making this app viable, and in general, I am pretty excited about it. I'm not even sure what I'm personally going to do with it, but several machine shops want to either consume its output, or OEM it directly.

They want it as a sales driver. Anyone can use it for free to design with... it saves the design on the client's computer. The file is XML and can be twisted into anything.

Or they can hit Order... and have someone build the part for them. Right now it only outputs DXF, but soon it will allow for a g-code module to be bolted on by a machining center to allow them to dial in the output to their machines.

It's a win for everybody... or so I think. What are the downsides that you see?

[Ron Dunn]

I think you've done quite a good job of programming, but there are a few points of detail needing attention:

1. The edit panel doesn't appear to resize when I change the panel dimensions.

2. When I change the grid size I get an uneven result. For example, changing from 0.1 to 2.0 gives an uneven measurement of the display area.

3. The rounded rectangle doesn't resize nicely in relation to the controls it contains. You can see that in the screenshot you posted.


Even with these fixes, there are some reasons why I wouldn't use it:

1. (May be wrong here) It looks like it will be tied to one vendor's panel service. Big mistake, but common in this industry.

2. It is really lacking in functionality compared to a some of the cheap front panel design products on the market, or even compared to simple drawing programs. I'm not holding this up as "the best", but take a look at Abacom's Front Panel Designer and its wizards and you'll see what I mean.

3. It is online only. That is the real deal-breaker for me, as I often do design work on the train, plane or other places where I don't have internet.

[kackermann]

Thanks for taking a look.

It's actually not tied to one vendor. There is a lot on the back end that you don't see. The app looks for a Site Config file on the hosting server, and the config looks roughly like this:


It allows processes to be defined as shop capabilities allow. Each process can specify a standard list of tools, or such things as offered colors for infilling, etc. It allows a list of materials to be specified, and options for the materials. The model supports a very accurate price quote engine that takes into account all material costs, machine operation times, and even tool wear.

The functionality at this point comes fast and furious. I am ready to put up a new version that has much better spline creation and fillet functions, very smart snaps that almost render a working grid meaningless, and a whole raft of other things.

I'm not aware of Abcom. The only Front Panel Designer software I'm aware of is from Front Panel Express. They seem to be the leader, so I use them as a benchmark, and my stuff already blows them away both functionally, and efficiency (I need a user manual!). They do have a few wizards, such as tapping that I'll blow away shortly.

Most of the effort so far has been in the development of the framework and standardized services and engines that the Bigger Picture is going to draw on. It looks like design software, but it is actually manufacturing software. I have a comprehensive polygon engine intended to preserve all higher-order shape definitions, such as quadradic and cubic Bezier's and ellipses through all boolean operations and out to g-code.

I'm dialing in all the best-practices for contouring (including pockets with islands), and sanity checking is underway. Soon, it will not allow a design to be submitted that cannot be safely and accurately manufactured. Lights-out manufacturing is the goal. That, and the delivery of friendly, powerful design software to the masses for free.

There is actually out-of-browser capabilities with the app, but to be honest, I haven't even tried it yet. It's very important to me that users do not feel captured by propritary data. I will be publishing the XML schema of the design file once it settles down. It's a bit verbose right now.

This is just one piece of the puzzle. My specialty is PCB design, and this framework will start migrating into other domains.

[tjskcnc]

Cloud computing (software accessed via the internet) IS the future of software. Operates the same way that applications in most companies already work - located on a server accessed through the LAN. Even Microsoft pays a small fortune annually to other companies to host their software this way, one company being a recent Motley Fool prime stock pick.
Wether or not your graphics program will survive this way or not, only time will tell. Hang in there and keep this forum updated with your progress. Who knows? maybe you will be the Bill Gates of online cnc and graphics programing.

[WayneHill]

I like it. You need to go 3D. Just think Google Sketchup CAM on Web...

[kackermann]

:-) Thanks.

I'm sure having a pile of money is great, and maybe I'll actually get to find out someday, but it's the engineering that motivates me right now.

The vision isn't new, but its execution can now be realized by an average Joe. No doubt this hints at being disruptive, but name any gain in efficiency that isn't. It's more empowering than disruptive. It should grow the pie, and not simply reslice it.

Someone who is graciously letting me use his machines and pick his brains while I dial in the g-code asked me a simple question: do you mean you can point your program at my shop and I'll get the orders?

Yes.

And how much would that cost me?

I haven't really thought about it. If I can make the process 10% more efficient than how you would manually process the order, then would a 10% commission be too much?
... so I get a sales vehicle for free, customers don't pay any more for their parts, and they have free access to design tools they never have to install, maintain, or upgrade.

It looks that way. It looks like I get paid, too. We can keep going vertical.

There are some very interesting possibilities that come about. Exposing design parameters to allow any shop to bid on any design, ensuring the customer is virtually certain to pay the lowest price, while at the same time allowing shops to pick off a set of designs that they aggregate into a single run, or to clear excess inventory or capacity by simply being the best price.

There are many other possibilities, too.

But I could have my head in the wrong place, too. I haven't thought about it yet.

[kackermann]

Thanks, and that's the hope.

I'm waiting for the client platform (Silverlight) to natively support 3D (it will, eventually). In the meantime, I'm going to make 2-1/2D the best it can.