Posts by dlanhart

Happy New Year all.

Is it possible to create function programs to reduce repeated code? There are many instances where we find the need to repeat functionality over and over again which could be reduced by implementing something resembling a function in traditional coding languages.

For example:

; Inside an arbitrary program...

CALL handle_signal(1049, 97)

CALL handle_signal(1050, 98)

CALL handle_signal(1051, 99)

CALL handle_signal(1052, 100)

CALL handle_signal(1053, 101)

CALL handle_signal(1054, 102)

CALL handle_signal(1055, 103)

CALL handle_signal(1056, 104)

; Handler program

.PROGRAM handlesignal(signal1, signal2)

IF SIG(signal1) THEN

SIGNAL signal2

ELSE

SIGNAL -signal2

END

.END

Quote from dlanhart

Thank you kwakisaki ! I actually found this shortly after posting, but I hope that this can be useful for others that are better at searching the forum than the manual.

Digging deeper, is there a way to add more condition complexity to the CASE statement? I have some programs that can be ran with a variety of commands so using less than/greater than operators are needed. For example:

Code

CASE command OF
VALUE 1:
  CALL prog1
VALUE 2:
  CALL prog2
VALUE 3:
  CALL prog3
VALUE >= 4:
  IF VALUE <= 10 THEN
    CALL prog4
  END
VALUE >= 11:
  IF VALUE <= 15 THEN
    CALL prog5
  END
VALUE >= 16:
  IF VALUE <= 20 THEN
    CALL prog6
  END 
END

Display More

I haven't tested this specific code yet but does this look like it would work or is 'VALUE n:' only able to represent VALUE = n?

I suppose for short spans of value I could comma separate, but using operators would be cleaner. Comma separated example:

CASE command OF
VALUE 1:
  CALL prog1
VALUE 2:
  CALL prog2
VALUE 3:
  CALL prog3
VALUE 4,5,6,7,8,9,10:
   CALL prog4
VALUE 11,12,13,14,15:
   CALL prog5
VALUE 16,17,18,19,20:
   CALL prog6
END

Quote from kwakisaki

Many ways and variations, that's what makes AS so flexible and powerful.
What you have is what I call bouncing IF's, where a value is being read to call a specific program.
After the call, if the value has been changed, then you run the risk of bouncing into another call based on the changed value.

If you are looking at just one call based on the value and then after completion of that call, a rescan then you would use a CASE (very similar to SWITCH but with plenty of advanced nuggets).
After the selected call, it would jump to the END of the CASE and not bounce into another call.
People find IF and CASE identical, and they are not in many ways but you can easily build and add process calls inside the CASE very easily.

Code

CASE command OF
	VALUE 200:
		CALL prog1;
	VALUE 235:
		CALL prog2;
	VALUE 240:
		CALL prog3;
	VALUE 241,242,243:; example of multiple values
		CALL prog4
	ANY :
		HALT; value does not equal any preceeding value
END; 

Display More

Look up CASE in the AS manual, I am sure you would prefer to use it, the above is a basic example.

More advanced techniques can include utilization of step 0 (parameters) a basic example is below which uses PROGRAM main, reads in a group of signals to create a BCD value and that value is read at the start of the CASE:

Code

.PROGRAM main();
  CALL prg_sel(BITS(2001,4));
.END

.PROGRAM prg_sel(.command);
  CASE .command OF
   VALUE 0:
    CALL prg0;
   VALUE 1:
    CALL prg1;
   VALUE 2:
    CALL prg2;
   VALUE 3,4:
    CALL prg3;
   ANY :
    HALT;
  END;
.END

.PROGRAM prg0();
  HOME;
.END

.PROGRAM prg1();
  HOME;
.END

.PROGRAM prg2();
  HOME;
.END

.PROGRAM prg3();
  HOME;
.END

Display More

There are further variations available, just a case (pun intended) of studying the AS language and exploring the possibilities.

Display More

Thank you kwakisaki ! I actually found this shortly after posting, but I hope that this can be useful for others that are better at searching the forum than the manual.

Digging deeper, is there a way to add more condition complexity to the CASE statement? I have some programs that can be ran with a variety of commands so using less than/greater than operators are needed. For example:

CASE command OF
VALUE 1:
  CALL prog1
VALUE 2:
  CALL prog2
VALUE 3:
  CALL prog3
VALUE >= 4:
  IF VALUE <= 10 THEN
    CALL prog4
  END
VALUE >= 11:
  IF VALUE <= 15 THEN
    CALL prog5
  END
VALUE >= 16:
  IF VALUE <= 20 THEN
    CALL prog6
  END 
END

I haven't tested this specific code yet but does this look like it would work or is 'VALUE n:' only able to represent VALUE = n?

Hi everyone, is there a better way to optimize how I am doing the following if conditionals to resemble something similar to JS where I can use switch statements or even if else statements? I haven't been able to find a clear way with the available tools in the AS language.

Example Code:

IF command==200 THEN

CALL prog1

END

;

IF command==235 THEN

CALL prog2

END

;

IF command==240 THEN

CALL prog3

END

;

IF command==300 THEN

CALL prog4

END

;

IF command==340 THEN

CALL prog5

END

;

IF command==350 THEN

CALL prog6

END

;

IF command==400 THEN

CALL prog7

END

;

Quote from kwakisaki

In my opinion and what I do, then yes 100%.
KTools and K-ide do not offer or provide me any advantages at all.

I can see your point for sure and that's where KTools and K-ide would prove beneficial.

However nothing like being able to bash out a reach study with all the data and simulated with cycle times and highlighting to design engineers that their cell design will not make prospective cycle times due to singularities and robot placement and EOAT tool collisions before they commit financially and present a solution to the clients.

Just my opinion though based on my bias.....

I think your point vs. mine is a perfect example of deciding which software to use.

As a preferred integrator of Kawasaki robots, the engineers over at Kawasaki do our reach studies for us before we land on a purchase decision which takes the need for simulation out of our work flow.

If we were to do the studies ourselves/had simulating needs, we would definitely choose KRoset.

Quote from kwakisaki

I couldn't say for sure, but I can hazard a logical guess.....
Revenue and the time they were developed vs where KRoset is now and where it is going.

KTools and K-ide were launched when KRoset did not have the functionality they provide.

Since then, KRoset revisions and development includes the functionalities now (not to the same level) but similar all the same.

I myself use nothing other than KRoset for all my needs and the stuff the others do not provide.

Pop over to the K-ide forum and ask the question......

Would you say KRoset is the best overall solution for Kawasaki programming? My team doesn't particularly like using simulators as they can be cumbersome slow programs and we prefer to do all motion work with the physical robot rather than simulating. This being said, I haven't had a chance to dive deep into KRoset as of yet.

Hi all,

I was wondering if anyone has information on why Kawasaki robotics doesn't promote KIDE on their robot software offerings page? From my understanding, KIDE was developed by a German branch of Kawasaki making it an 'official' Kawasaki product, so I am confused as to why it is not marketed more. In fact, the only reason I was able to find the software was by browsing this forum trying to find a way to create a plugin for VSCode that contains the functionality of KIDE.

The KIDE software is absolutely incredible and has significantly increased our development process as it essentially merges functionality of KRterm, KTOOLS, and the teach pendant. I honestly don't see why anyone would use any other offering currently available.

I hope this post can not only answer my question as to why KIDE is not more widely known/pushed, but also serve as a marketing opportunity for the software so more developers can discover it for themselves.

Here is the post that describes how to setup KIDE as well as other development environments for Kawasaki robots: Click Here

Quote from dlanhart

Thank you kwakisaki and ShAdOwDrAgOnS for the helpful response.

I see the menu now and was able to locate the Collision Detection manual, that was separate from the main AS language manual.

Now regarding how to enable collision detection, here is my current code that is NOT working as when we try to manually collide with the robot, it is not stopping or giving any indication that the collision detection is working:

.PROGRAM main() #0

BITS 65,16 = 0

COLRJON

COLRON

COLR 0, 0, 0, 0, 0, 0

COLRJ 0, 0, 0, 0, 0, 0

SPEED 1 ALWAYS

ACCEL 1 ALWAYS

DECEL 1 ALWAYS

ACCURACY 1

continue = 1010

z_clear_height = -327

HERE current_pos

DECOMPOSE current_pos_arr[0] = current_pos

LDEPART z_clear_height-current_pos_arr[2]

JAPPRO pickpos,100

COLRJOFF

COLROFF

SPEED 100 ALWAYS

ACCEL 80 ALWAYS

DECEL 80 ALWAYS

main_loop:

Any help on this would be greatly appriciated, thanks!

Display More

I was able to solve this myself, the solution was to change the values of COLR and COLRJ to a non-zero number. A value of zero will be ignored by the detection systems for the respective joint.

Thank you kwakisaki and ShAdOwDrAgOnS for the helpful response.

I see the menu now and was able to locate the Collision Detection manual, that was separate from the main AS language manual.

Now regarding how to enable collision detection, here is my current code that is NOT working as when we try to manually collide with the robot, it is not stopping or giving any indication that the collision detection is working:

.PROGRAM main() #0

BITS 65,16 = 0

COLRJON

COLRON

COLR 0, 0, 0, 0, 0, 0

COLRJ 0, 0, 0, 0, 0, 0

SPEED 1 ALWAYS

ACCEL 1 ALWAYS

DECEL 1 ALWAYS

ACCURACY 1

continue = 1010

z_clear_height = -327

HERE current_pos

DECOMPOSE current_pos_arr[0] = current_pos

LDEPART z_clear_height-current_pos_arr[2]

JAPPRO pickpos,100

COLRJOFF

COLROFF

SPEED 100 ALWAYS

ACCEL 80 ALWAYS

DECEL 80 ALWAYS

main_loop:

Any help on this would be greatly appriciated, thanks!

Hi all,

After looking through the AS manual I have not been able to locate where in the literature it describes how to change the limits on joint collision detection.

For context, we are trying to 'dummy proof' a system where if an engineer was to jog the robot manually to a possible collision area, then put the machine in auto, which would automatically send the robot to a 'home' position, the robot moves slowly to that safe position before ramping the speed back up to production. We have the speed setting in place and working, but aren't sure how to make the collision detection as sensitive as possible?

Thanks!

I have come up with a solution, that fits my team's specific needs, and this was the work around:

1. Our autostart.pc program was not declared correctly, this was the correct syntax:

.PROGRAM autostart.pc()
.END

; was previously...

.PROGRAM autostart()
.END

2. We needed our MAIN program to run on power up so we added to autostart.pc:

.PROGRAM autostart.pc()
  WAIT SWITCH(POWER )
  MC EXECUTE main
  10
  	; other functionality
  GOTO 10
.END

Thank you kwakisaki for the input, you are correct with how I outlined the conditions for returning to MAIN using cycle start as the indicator in my previous thread. After digging into the weeds more for this project, the solution above works best for what we are needing now.

Hello all, I was wondering if there is a way to auto select a program in the event of a power loss/restore? For example, I am in my PICK program when power loss occurs and I want to get back to my MAIN program when power is restored. Thanks!

Everything is working now as intended, thanks for the assistance!

This seems to work perfectly for my application.

The only issue, my application specifically, is that I am already using autostart 1 as an end of arm clamping process and is conflicting with the logic for auto cycle. Is there a way to fit both in the same program or should I separate them between two autostart programs?

Thanks!!

.PROGRAM autostart() #10
10
.cycle.on = FALSE
  WHILE TRUE DO
    IF SWITCH(CS) AND .cycle.on == FALSE THEN
      .cycle.on = TRUE
    END
    IF NOT SWITCH(CS) AND .cycle.on == TRUE THEN
      MC PRIME main
      .cycle.on = FALSE
    END
  END
  IF SIG(1011) THEN
    SIGNAL -50
    SIGNAL 49
  ELSE
    SIGNAL -49
    SIGNAL 50
  END
  GOTO 10
.END

Hi all,

I am currently trying to implement recovery to my robot and I believe that I have everything (code wise) in the main program solidified, however when in cycle start and in a program other than main I cannot seem to find a way to return to main either before restoring cycle start or after. Any help on this would be great, thanks!

In the example, I would like to go from .pick to .main when cycle start is turned off so that the z_clear_height recovery process is triggered before jumping back into pick when cycle start is turned on.

code:

.PROGRAM main()
; set current position to a point
; z position at mold that is clear 'clear z height'
  z_clear_height = -456
  HERE current_pos
  DECOMPOSE current_pos_z[0] = current_pos
  LDEPART z_clear_height-current_pos_z[2]
main_loop:
blah blah blah


.PROGRAM pick() #1075
  BITS 65,16 = 0
  ACCURACY 100
  JMOVE apprpickpos
  ACCURACY 1
  LMOVE offset1+pickpos
  BREAK
  BITS 65,16 = command
  SWAIT continue
  LMOVE offset1+partdetectpos
  BREAK
  BITS 65,16 = command+50
  SWAIT continue
  position = 0
  ACCURACY 100
  JMOVE apprpickpos
  BITS 65,16 = command+1000
  RETURN
.END

Quote from Motouser

I use a modified version of Notepad++. It establish a TCP/IP connection.

To clarify, are you able to save/load to and from the controller using Notepad++? If so, are you able to share how to establish this connection? I am also looking to do something similar to OP, but with creating an extension for Visual Studio Code that integrates some functionality (like load/save) form kTerm.