Paid (!) Help Wanted for MPCNC pick and place

Software / Firmware General
1

Dear MPCNC-Community,

I reach out to you guys as we are looking to automate certain aspects of our production line. That is we are a producer of plastic candle cups in various sizes for our local clients here in Poland.

We currently place these cups by hand, and since these are produced and sold in the millions you can image the strain it puts on our workers.

Several attempts to automate with industry standard equipment failed to to cost concerns. Also, we are quite far away from a support structure and service visits, changes in programs with proprietary service technicians are not an option for us.

Therefore I was searching for a solution based on open-source technology and based on something that is already there. And this is how I found the MPCNC.

I hope to take the MPCNC to a size of 1400x900mm (x&Y) and (350-500mm) in Z (so a standard Europalet 1200x800 fits under it).
The precision and repeatabilty is ok at ±1,5mm. So not more than what a human can resonably achieve over longer intervals and given the cycle time.
The travel speed required are between 50-100mm/s. Of course faster is better, but let’s stand from there.
The Gantry can be belt driven (as is the MPCNC) or screw driven.
The payload is virtually none. Each cup weights in at not more than 25gr of PP. If we take 16-20 pcs really just the matter of the gripper is what we carry around.
The gripper can be linear actuator where we print most of the contact surfaces (I will explain more if needed).

So what I really need is a machine that picks up a set of cups on one end of the table and puts them away on the other end until the array of cups is completed. Then it can wait (accustic signal) until the operator has cleared the area below and restarted the sequence with the push of a button.

The challenge in this project is to implement the gripper, the logic that decides whether a set of cups is ready to be picked up, the push button for program reset/repeat, the implementation of possible closed loop steppers.

Now I know that many of you have spent years to acquire the skillset to pull something like this of, and certainly you spent the other buck or two. Now this is a chance to earn some of that sweat 'n tears money back as I don’t expect you to do this for free. It’ll certainly make cool content for social media.

To be clear I am paying you for your expertise and knowledge not for the MPCNC parts, which we will print ourself or order from V1 as much as possible.

Please let me know how I can contact you. And as we are here in Poland, European helpers are preferred.

4 Likes
2

That sounds like a really neat project. I thunk you are right that the gripper is where the biggest challenge will come in. Marlin (and grbl) aren’t very flexible when it comes to decisions. They just follow the instructions they are given. So you’ll need some software to decide which cup to pick up and where to put it.

Sounds like you need a good mechanical engineer and a decent software engineer. If you want it to be completely automatic, you’d need a great software engineer. I hope you get some help and share what you’ve found.

3

The gripper could be something as easy as a small rubber cup with a vacuum.

2 Likes
4

So lets see if i understand the order of operation.

1 arrange cups in specific position
2 remove from work area
3 operator pours material
4 moves cups out of work area

It seems this process is over simplified and you might not be best served by the mpcnc. I believe you are looking for a 5 or 6 axis robotic arm.

However if you are set on using a mpcnc it is not impossible.

You will need a micro controller to command the logic and tell the mpcnc which program to run, if you want it to run full auto.
Otherwise you could just save 2 programs (1 for placing and 1 from removing) and have your operator select the program

5

That’s a little gloomy. It is true that an external microcontroller or software could make it a lot easier, and possibly be needed, but it doesn’t exclude the MPCNC.

6

Especially since you could, for this specific need, do something like a LR2, but instead of a router plate, have an array of pick’n’place heads that can place a row of cups at a time. Put in a tray, place the cups, pull the tray. Or even get really crazy, and change the Y-axis to a belt feed, and have a worker feeding empty trays that get fed under the P’n’P heads, and another worker pulling full trays off the other end. Technically, that even frees up the X and Z axis for use with the P’n’P heads.

edit: Disclaimer: I am not an engineer, I’m high-functioning autistic (AKA Asperger’s) and I watch How It’s Made as background noise…

1 Like
7

Thank you for your posts so far. I will explain further.

IMG_20200224_162521
IMG_20200224_162521.jpg2736Ă—3648 3.32 MB

This is how these cups are coming out of the machine. I plan to prolong this rail across the entire length of the table (short side 800mm-900mm).

8

IMG_20200224_162513
IMG_20200224_162513.jpg2736Ă—3648 1.81 MB

This is the array our employees have to place them in. In this case it is one row of 15 PCs then 16pcs then 15pcs (to a total of 27 Rows). At 27 repetitions the machine would need to sound a signal.

Then the worker comes to slide the card board tray over (roller table!) to the next station (manual palletization) before placing a new tray into the machine and pressing repeat.

This is a semi automated work flow but that is 100% acceptable.

As for the gripper:

Suction cups would be great but this would require some sort centering setup to grab all 16/15 cups individually. This could get expensive really fast and we would need to figure out precisely how to make cups stand in the same position all the time to avoid collisions.

Therefore I think it best to grab them from the outside with a cutout shape of 16 holes and just move one actuator to grab them. This makes it faiy easy to align them relative to the gripper.

1 Like
9

I dont mean to sound gloomy, just making sure they understand that this sort of use is (imo) beyond the usual and intended use. Though it would be simple (again imo) to adapt it for this use. Id be glad to help anyway i can but i can only lend theories as i havent built one yet

1 Like
10

Depending on the degree of sterility needed, it might work better to have the gripper grab from the inside of the cups. Then it can place them closer to the previous row/tray edge. Think a pair of chopsticks with rubberized/foam ends.

11

Since you are grabbing a whole row at once, and the fact that there is no need for this machine to do anything else, the MPCNC might need to have more than half the machine replaced. The Low rider seems like it might be a better fit.

The software seems way easier now though. Since it is always the exact same 27 operations. There will be huge messes made, because there will be some slop, and some cups will fall over, and then huge swaths of cups will fall over. The machine will not be able to know when there has been a mistake.

How would you pick up either 15 or 16 cups? Maybe when you want 15, the 16th hole is placed past the end of the row?

1 Like
12

I get it. It is definitely not an “off the shelf” task for the mpcnc. I am always excited to see it do something new though. It wasn’t meant for plasma cutting, or aluminum milling, or laser engraving either.

3 Likes
13

I think you are asking for a open source palletising machine.
Have you contacted zechetti?? They service a large industry

Is the closet video they have to what you are asking for i think

1 Like
14

The “grip head” could be a mpcnc and the drop head could be a lowrider but both would require significant alterations to the design imo

15

If the trays are just for transfer, build two LR2s. One to make custom transfer trays with more exact cutouts, and the other to do the PnP. A 6mm HDPE sheet with chamfered holes in it would work wonders. All depends on what your manual palletization process is…

1 Like
16

After seeing the pictures, I think you’d be better served by having the cups flow into some type of long corral where they could be dropped out the bottom. Have the box move under the tray after each drop and have something counting the number of cups that went into the corral to know when it’s full for that row. The same conveyor that’s moving it after each drop could then feed it onto the rollers sending it down the line. Your employee would just have to place the box at the start of each set.

17

Sterility is not a factor. Grabbing from the inside would be OK, but I worry about the exact positioning requires for that sort of thing.

18

These cups cannot self propel by being pushed from one side beyond a certain point. After like 20 to 30 PCs in a row we can see deformation in them, hence we’d need to aux-propell them with a conveyor is similar

19

Ah ok. i think i under stand what you would want, and how it could be accomplished without the use of a logic controller however it would take a minute to program the gcode as it would have to be made almost completely by hand

20

This solution is easy as this could be achieved with wait for user command in marlin

Id like to try to help, however im in the us. Also for the sake of honesty i have yet to build a mpcnc myself. Feel free to send me a pm on this forum and ill send you my contact information.