What's new
What's new

A simple robot thats easy to program.

Turbo17

Plastic
Joined
Aug 7, 2010
Location
Southern California
Hey guys.

I was wondering about a robot for loading and unloading of parts into a lathe.
Do they make a simple robot arm that has a simple programing language?
It only needs 3 or 4 axis to swivel and articulate a pre-programmed set of movements.
Also a "HAND" that closes and opens to hold material.

Does such a thing exist?
Or are they all complicated?

PS: It seems to me that a simple load/unload robot arm could be programmed with Gcode, but im not sure.
Do you guys have any suggestions?

What say you?
 
None exists - afaik.
Baxter is perhaps nearest at 20k$ or so, plus n, x, y, z, ...
Grippers, actuators, guardings or rails, certs, integration.

And industrial robot is about 40k€ wholesale, lathe, 3 kG end-effector mass, and 3x when working.
About 100k and up.

I may be working on this ....
Value is important.
If You think a 5k all-in, this is unlikely short term, due to industrial legislation.

If You think 30k total is feasible, then well ....
You know what to do.
 
All robots are easy to program.
They just speak a different language, early days many did g-code.
You walk and talk G00, G03, G54, G42, G81 which makes no sense to a common man.
They are just different but still machine tools.
Bob
 
Sawyer Collaborative Robots for Industrial Automation | Rethink Robotics

Sawyer is the next version of Baxter. It only has one arm and is geared towards loading machines and things like that. Baxter is more suited for packaging type tasks. They are both programmed by physically moving it through the motions you want it to perform.

It's amazing how much that thing shakes in the video. Watch closely when the gripper tries to pick up the part from the pallet and again when it tries to load it into the chuck.
 
A real industrial 6 axis robot is only $30k new for a one-off with IP67 protection, which you NEED for a machine tender.

All robots are falling down easy to program for simple pick and place, especially if they are owner operated so you can just jog them out of screw-ups instead of having to code error handling routines.

Why are we talking about Baxters (a complete joke) and Universal (50-100% overpriced)?

This application is easy for any robot. Pick one and get at it.
 
A real industrial 6 axis robot is only $30k new for a one-off with IP67 protection, which you NEED for a machine tender.

All robots are falling down easy to program for simple pick and place, especially if they are owner operated so you can just jog them out of screw-ups instead of having to code error handling routines.

Why are we talking about Baxters (a complete joke) and Universal (50-100% overpriced)?

This application is easy for any robot. Pick one and get at it.

This conversation is actually ongoing in a couple of different threads right now.

With a real, industrial, 6-axis robot, the price of the robot (and controller) are a small part of the total cost of the system. Opinions vary, but the prevailing thought between the automation guys (Tonytn26, Motionguru, and I'll include myself, even though I'm not in their league) is that the total system will cost from 2X (if you do this all day, every day, and have existing designs for the mechanical and controls systems) to as much as 4-5X (for new projects, clean sheet designs, etc.) the cost of the robot, to get a new 6-axis robot integrated properly.

So a $35,000 robot turns into a $100,000 automation cell, very, very quickly.
 
Yes, but how is a universal robot any different? No more than $5k of the integration cost is the cage. What else can you skip with an overpriced collaborative robot?

I don't disagree that the cost of a well engineered cell is between twice and three times that of just the robot, including labor. But what costs of those does a collaborative robot, which has to be programmed a good bit more carefully to be safe, save?
 
Collaborative robots have serious speed limitations due to the collaborative aspects. Inertia and all that..... Most folks doing industrial stuff won't be keen to taking a very large cut in productivity speed to reduce out a person, it is just not economically feasible in most industries.
 
I haven't had any robot experience at all.

As the business here grows, I'm hoping that I can use one of these collaborative types to limit hiring people. I don't want to set up something to run the same part for years and years. What I'm hoping I'll be able to do is set the thing up to change a couple hundred parts in the lathe, while I set up the next job in another machine, as an example. If it worked 2/3rds the speed of a human, it would still be better than having to run over and change parts myself every 3 minutes, yet it could be monitored without too much trouble. Maybe this isn't practical, or maybe it isn't practical now but maybe it can be as the robots continue to improve.

If there's no parts to run tomorrow with no real quantity, the robot can sit on the shelf for a couple of days easier than figuring out what to do with an employee for a couple of days. It might not be the total solution, but it sure seems like if something could just tend a machine now and then, it would sure help keep the labor down in a smallish job shop environment.
 
I haven't had any robot experience at all.

As the business here grows, I'm hoping that I can use one of these collaborative types to limit hiring people. I don't want to set up something to run the same part for years and years. What I'm hoping I'll be able to do is set the thing up to change a couple hundred parts in the lathe, while I set up the next job in another machine, as an example. If it worked 2/3rds the speed of a human, it would still be better than having to run over and change parts myself every 3 minutes, yet it could be monitored without too much trouble. Maybe this isn't practical, or maybe it isn't practical now but maybe it can be as the robots continue to improve.

If there's no parts to run tomorrow with no real quantity, the robot can sit on the shelf for a couple of days easier than figuring out what to do with an employee for a couple of days. It might not be the total solution, but it sure seems like if something could just tend a machine now and then, it would sure help keep the labor down in a smallish job shop environment.

If you've ever done any computer programming or have some knowledge of logic/flow control concepts(I mean really really basic stuff like loops, if/then, etc.)then you can do a lot very easily with the UR. It also helps if you can wire up relays or pneumatics and use a multimeter. Most of the programming is done by physically moving the robot to various waypoints. Pretty easy. We're in a similar situation where we have products that sell in spurts, then a trickle. The robot and lights out machining lets us buffer those spurts without having to hire and layoff people all the time or deal with temps. So for us a fast cycle time isn't all that important either. We can load up enough stock to let it run for twelve hours when we need it. Some nights we don't run it at all. I actually think robots are ideal for small (even one-man) shops.
 
Here's a video someone posted to Facebook yesterday in a machinist group, showing the exact set up that I'm picturing in my head:

robot - YouTube

I have no details on his set up yet, but he's posted other videos of him running the same machine without the robot. Looks like he just shoves it out of the way when he doesn't need it. I believe he's in the UK...
 
So think how much money you could save just automating the door, for robot or human! Cycle finish pops the door open while robot or operator is prepping next part. Time savings either way.
 








 
Back
Top