I think I told you about my limited experience with FANUC programing with CNC machines. The programming was automatically done by CAD programming. All that needed to be entered was coding for the material to be machined, the cutter material (HSS, carbide, etc.), coolant type, etc. With this info it figured things like cutting speeds and depths of cut on its own. I wonder if there isn't a similar program for FANUC Robots where you enter package size, weight, location and layout of parts to picked up, etc.

Is the pitch, roll and yaw controlled with it's own programming parameters or are they programmed with combinations of X, Y and Z locations and distances? My experience was when there where only 3 axial coordinates were used and not the 5 and 6 they use now which I think are X, Y, Z, A, B and C. But even then, X, Y and Z were for the table movement and A, B and C were for the cutter head orientation. I imagine that's quite a bit different than a free moving arm locating itself?

I'd be interested to know something about controlling the sensitivity of the pick up mechanism for different types of lifting, carboard box vs. a solid metal block.

 

It is X, Y, Z,P, W, R.

I also believe that there is a program that we can purchase to integrate this a little easier. However, someone really needs to know the KAREL portion of it in case of catastrophic failure. There are certain things that MUST be done with these robots, or you'll be starting all over again. I need to research it deeper this week, but this is the program that the instructor referenced a couple of times.

 

which one we talking about? ive stopped reading quite a few already LOL

 

Power steering guy....

 

Interesting. From the videos it appears everything is picked up by suction? Is that what you're learning? Also, where you work do they have the machine that makes product just appear out of nowhere like in the videos? I want that machine!

 

ah yeah the train wreck... couldnt not read it just to see where it ended up but i refused to comment lol

also geez what a pain in the azz changing the rear shocks was on the escape. Both upper nuts had to be cut off (despite being inside the car and completely rust free. Then the driver lower nut wouldnt budge with the impact so i put a socket wrench with a cheater bar and ended up breaking the tab off the bolt that holds it in place... sooo took off the tire and used a breaker bar to brace the bolt from spinning and then back to the socket wrench with cheater bar to SLOWLY get that dang nut off! Only positive was it went back together super easy and now the wife has new rear shocks. The ones on it were original so 244,000 miles old lol. Air filter is from 2015 whoops! so going to buy a new one tomorrow, doesn't look too bad tho. Then topping off the power steering. I'll have a video for you guys tomorrow to look at if ya dont mind. Cant tell if the blend door is moving as far as it needs to be. Trying to find ANYTHING that will get her heat so I dont have to replace the heater core... but I think thats all i'm left with.

The shocks started cz while getting my lifetime alignment done at Firestone they mentioned the rear driver camber was slightly off but theres no adjustment to it so theres something else wrong. Tech mentioned shocks and then i realized we haven't touched them so they are at least 100,000 miles old. After taking them off they were definitely pretty soft, especially compare to the new ones. Also I then started to think about it and I don't see how that would really affect my camber so any thoughts on whats causing the negative camber in the rear? Can post the alignment printout if it would help

Dang that turned into a long post

 

No, we only learned how to measure for, and align the machine with the end effector. AFAIK, Fanuc does not make any end effectors, and you have to outsources that. The particular machine I went and looked at in operation had a suction unit made by SRS.

The shocks on the Escape are a PITA. The struts are even worse. Every damn one of them (shocks) that I've done wound up getting cut, just like you did. The right rear shock on my 2011 started squeaking tonight.... I think the same thing is going on with this one as my 06. The rubber sounds like it's gone hard.

 

There are certain coordinates in the programming that, after initial setup, make the robot move as a unit instead of joint (individual motors) mode. World mode (which IMO should have been called Global) and Tool mode allow it to move as a unit when jogging. Really and truly, it's all in how you program it. Oh, and table movement... I'm not sure what you mean.

Here are some pics I took of one of the programs I did to stack three 50mm blocks.





The possibilities are endless here, and to be quite honest, we really didn't delve into this much, because like I was saying... the robot will not come with the tooling (end effector) you need. All we really talked about was how to set up the weight of the end effector, and also the weight product within the program... and it automatically calculates max speed from there. If I think that is too fast and unsafe, I can slow it down too.

 

Wow, you're fortunate to get all of this.

First, by table I'm referring to a machining center. Think of a drill press where the drill is in one location. If the piece being worked on has multiple holes to be drill in to it, the entire table the piece is clamped to moves to the location of the holes. That's the X, Y and Z although for a drill press operation the head may have travel just like a typical drill press but the table can lift in to the drill bit also. A, B and C are angles the cutting head takes. For example, a 45° chamfer on an edge, the cutting head will rotate to the 45 and the entire table will move the piece for the cut. Just like the X and Y axis's on a Bridgeport Milling machine. Translating A, B and C for a cutting machine to the P, R and Y of a robot I imagine is how a cutting machine used for CNC porting of a head would translate to a robot. A, B, and C are primarily sine and cosine identities used for locating in a 3d space.

I've put KAREL on my to do list. A list I keep of things I run across and want to google when I have the time.

World vs. Global; World is an odd descriptive word. I had an automated production line explained to me years ago and they used similar odd words for processes. The entire line was called a universe. Each station was a planet with a name. Each process that moved parts between planets was interplanetary with its own name which was also related to how the parts moved from one station to another; conveyor, gravity power roller tables, robotic arm, etc. I assumed this was done to use somewhat familiar and less intimidating words\terms. Using them avoids technical terms. When you think about it most of our language and communication is slang instead of proper nouns or terms.

Yea, I'm mostly interested in how the robots actually move parts. The moving of the robotic arms and positioning them seems to be almost universal. That's something they all have to do but the actual handling of the parts or whatever seems to be the real focus of the technology and the future. I can see how a high school science fair had a challenge of a robot being able to pick up and package an egg then picking up a 20 pound cube of steel then ending with something like relocating a broom handle through a maze of potential interference. I wonder if there are universal, self sensing pick up devices? I'm guessing the more universal the pickup device is, the more valuable it is to end users?

A last question. How is "fine" defined in the programming. To me it suggests "acceptable" but that also suggests a lack of precision I'd expect from numerical controls?? Maybe it's just up there with "world"?

 

With a robot, you would want the table to be stable. I don't see any benefit in moving the table around if it is in the work area (within reach without hitting singularity) of the robot.

As far as "self sensing", there is some of that in the market already. I was reading recently about a company that made an end effector to pick up pastries with a robot. Not each pastry is identical, so IMO... it would need something to sense how tight to grasp. I'll have to look for that article again and re-read it. Also, there are Cognex, Keyence, Fuji vision systems that you can integrate into these. The robot I was using had a Fuji on it, but that was not covered at all in my class.

As far as the "FINE" is concerned. The alternative to the choice is a CNT (continuous) tag. We didn't get REALLY deep into this, as it is more of an advanced class deal. I will explain it to the best of my knowledge. There is a lot of calculus equations, trig, etc. that go into the most efficient way the robot can get from point A to point B. The most efficient way is NOT a straight line, as the human brain would think. The robot calculates the most efficient use of power, time, and the least amount of wear on the motor, and will use that under the CNT tag, which is followed by a number of 0-100. CNT0 is equal to the FINE, while the CNT100 pretty much tells the robot to do its thing, and be efficient. It will make shortcuts, round corners, etc. Not something you'd want to do with a welding tip. The 0-100 a percentage, as I recall. So yeah.... if you've got anything remotely in the path of the robot that could be hit, CNT may not be for you. Also, you will see J and L in front. The J is for Joint, the L is for Linear. You always want to use a linear motion on approach and retreat from your destination points.