RE: Home shop CIM (2)

Eric wrote:

>My goal at the time was to build a CIM cell that would
>produce alphabet blocks for kids with their names on it.
>I wanted the robot arm to pick up a 2" block, put it in
>the CNC router, flip it over as needed, and eventually
>hand it to the kids on the other side of the Plexiglas partition.

This is exactly the type of ad-hoc workcell capability that I am
interested it!

>I did not realize (at the time) just how complex this
>project would be. Now, I am focused on the CNC building part.

Yeah, the world seems to be designed to lure a person in over their
heads thereby tricking them to learn some tough things! The funny
thing is that we fall for this trick every time...

>But, your idea intrigues me and I still have all the parts.
>Please do keep us/me informed.

Well I am at the point where I am trying to locate people who are
intrigued with this idea so that they can help 'flesh it out'. After
looking around the net a bit I have concluded that a number of the
members of this CAD_CAM_EDM_DRO list have the right kind of background
for this type of project.



Chris said:

>Hey thats a real NEAT idea.........perhaps you need some help
>programing or some ideas if you get stuck?

Open source development thrives on open collaboration and it would be
great if you and others were interested enough in this project to help
develop it. Are you in to Java at all? If not I have a version of
this free online Java beginner's course based on the TINI/TStik that I
can make available to people on this list:

http://www.javadevices.org/dtcourse/jitv_syllabus.html



Dave said:

>>Often it is easier to implement a bit of electrical
>>handshaking between machines as opposed to complicate
>>I/O networking.
>>
>This part makes the most sense. run the machine tool till
>done. send an output done signal and wait. the
>arm/carrier/part loader then gets a start signal and then
>it starts to run it's program till done, then it waits
>and the machining starts. simple electrical handshaking.
>no real need for networking.

I think that it is much easier to just attach everything to the shop
network and let the software take care of handling the message
traffic. With this type of a system, machine interactions can be
changed at the press of a button without getting out the electronics
tools and wire in order to put together ad-hoc hardwired interfaces.

Carburetors use to be considered 'easier' than electronic fuel
injection at one point in time too. The thing to ponder, though, is
why a system that is an order of magnitude more complex than the
simpler system has almost completely replaced the simpler system? To
me it is so a person can just turn a key and the car just starts
without needing to think about it.

>I see that programming would almost have to be dependant on
>materials more than anything else. unless you have some vast
>supply of all the rought stock for all the possible projects,
>you would be limited to starting off with some pre-determined
>blank and machine the heck out of it.

Yes, with industrial CIM one designs the raw material's configuration,
holding fixtures, machine cuts, etc. all at the same time to
accomodate each other. Instead of designing the system to manufacture
all possible part types one designs the system to accommodate a family
of part types.

But you are right, home shop CIM will probably need to use a different
strategy than this which is the type of thing we need to discover.


Ted