Some questions, first.  Understanding your system configuration will help provide answers.

(1) is this a standard industrial robot, or a home-grown device?
(2) Do you want to control the robot's motors directly using the RS232?
(3) **OR** do you want to send command strings to the robot controller via RS232 and have the robot react to the command strings?

TygerDawg

Do you have a robot controller ?
Does it have a command set which control can be used to issue commands to the robot ?

If so I would do two programs.
The first would be a lower level program issuing commands to the controller and listening for responses. It would translate seemingly english commands like PICK AT LOCATION 1 to the machine language the controller. PICK AT LOCATION 1 would tell the controller to guide the robot from a safe location though a predefined path, pick an object at location 1 and return though that path to the original safe location. It would also translate responses from the controller (like &*## means ERROR: AXIS 1 REACHED LIMIT).

The second program would generate the logic required to control the robot so that it meets your process requirements. It would issue commands PICK AT LOCATION 1 and listen for response (DONE, for example) and be a higher level program with a user interface.

If you are trying to plug in the robot directly to a computer, I think this would be tough to do in VB. There are a bunch of issues here like power regulation, position tracking, coordination of axes, e-stop regulation....

Cheers

about the control robot!! i want to control the robot's motors directly using the RS232, this is a standard industrial robot (mitsubishi),... thank you!

I wrote an application (DOS based but with a real graphic interface) which can be connected to :
     STAUBLI UNIMATION robots
     ADEPT Robots
     FANUC Robots

It controls Pick and Place application, but also process applications like deburring, welding, glueing.

I cannot say more, I don't want to advertise an obsolete product.

Alain Riaud

you won't get enough data through a RS-232 to control even one motor directly. It is not a suitable interface for that job.

Jay Maechtlen
http://home.covad.net/~jmaechtlen/

Jay, I disagree.

I use RS232 to control, in real time, 6 axis robots on fairly fast (300 mmps) trajectories. I have more than hundred robots running that way, making shoes or reparing hydropower turbines.

Com speed is 19200 bds, refresh rate is 16 ms which is enough for 16 ms  robot controllers.

Jay, I agree, no way to control directly motors on RS 232, that's why a robot controller is needed for that task.

If you want to control directly motors, you have to use a fast industrial LAN. I have done it on a 32 axis screw lathe
with 6 ms refresh rate on movement and less than 1 ms refresh rate on motor control.

Regards

ok, I stand semi-corrected?
If not a robot controller, at least some kind of axis controller, right?

I'm curious: what do you mean "control in real-time"?
Are you interpreting an NC part-program (or the equivalent) and feeding coordinates to the controllers, or what?

Jay Maechtlen
http://home.covad.net/~jmaechtlen/

Jay,

To be more accurate.

The robotic application I'm speaking about is sort of machining application, welding , grindig or glueing.

By control in real time I mean, at the PC level, every 16 ms,new coordinates are sent to the robot according on its actual position received every 16 ms throuh the same channel.

That means all trajectory planning is made at the PC level and this planned trajectory can be altered every 16 ms if someting changes in the environment.

For that, I do need a robot controller.

For the NC machine, the one I developped was based on a LAN for other reasons.
It was a two tiers architecture:
One PC used as a MMI, interpering and sending NC commands to
a real time (6ms) PC feeding the motors servo through a CAN bus. The refreshing rate (6 ms) was limited by the CAN bus and the servo, the PC was able of 2 ms.

Jay, as it seems it's a conversation between the two of us, I suggest, if you wish, we carry on privately, unless some other members shows up.

Keep it public if you please.  I've done lots and lots of Staubli V_TRAJSIG work. I'm listening with great interest.  

TygerDawg

Tygerdawg,

Ok I will keep it public.

As you are familiar with V_TRAJSIG, you should know Jacques Vuachet who left Staubli recently.

We worked together on the ancestor of V_TRAJSIG in the mid
80's, then I left Staubli. I then worked on my own version of V_TRAJSIG, written in C/C++, running on PC, with a graphic MMI, embedded in what I call a "robot cell controller".

This "robot cell controller" based on a PC can communicate through RS232 link with FANUC Karel robots and UNIMATION/  STAUBLI or ADEPT ValII and V+ robots.No need to write again your application in a new language.

Unfortunately, I have not yet connected it to CS8 controllers.



 

Yes I know Jacques very well.  It's a very small world out there in Robot Land.  I hope that he's doing well "Apres-Staubli".

What you have done is very impressive.  I always thought V_TRAJ was extremely powerful, but very complex.  Most of the US customers could never understand how to use it.

Ever since I was also "Apres-Staubli" I am the US reseller of this product:
www.robotworks-eu.com
I always thought it was a good match for something like V_TRAJ.  But the market is not ready for it yet I think.

TygerDawg

Alainriaud

Your new version of  V_TRAJSIG sounds interesting. Is it ready to sell or can you give information of the current status?

Jonas,

It's not a new version of V_TRAJSIG, but a complete environment including V_TRAJSIG looking capabilities .

This is not a real product (advertised, documented ..) but a bunch of softwares I used to build process applications for the last ten years. More than 100 robots controlled by this software.

This software is not for sale as is, but I'm ready to build, or help anybody to build an application with it.

One important thing you should know, it's based on DOS, even if it looks like based on Windows. I'm working on a Windows version which is far from completion.

I don't want to promote this software any longer on this forum, it's not the right place. Please contact me in private.

Alain Riaud

Alain,

I appreciate your information of your current work. Since I don’t know how to contact you privately I hope you give me a mail at   Jonas@itproffsab.se

Jonas Stahl