Sensor for counting...

[CroCop]

Can you zip them?

 

[darrenj]

yep..i am doing that now..but curious how i can get them dispalyed in my post..I have done it before but cannot for the life of me figure out how!!

Zipped files will be here in less than 5

 

[darrenj]

here they are...

 

[darrenj]

bags1.jpg

Typical tight tag bag

bags2.jpg

typical untagged bag..(Broken bag)

bags3.jpg

Not quite typical bag..(My kids already had one "pouch" of milk..But it gives you a general idea of what it looks like..

pic5.jpg

Typical Pouch of milk..There are 3 of these per bag..each bag has 4 liters of milk..


BTW not the brand name of milk i work on...The names have been changed to protect the innocent!!

 

[CroCop]

Do they stack against eachother?
Like right next to eachother with no break inbetween on the line?

I still think you could get one to sense the conveyor, and let the change of not being on the conveyor trip the counter.

 

[darrenj]

there is usuly a gap..however if you have a broken bag then they will run together..most times you have a 1/4" gap or so..

 

[CroCop]

So there is no gaurunteed break between the bags?
That's tough.
Vision might be your only way to count accurately then.

 

[TDMartin]

Look at the Checker product from Cognex. It's new, somewhere in between a photosensor and vision system. It's marketed for applications like this, that photosensor's can't handle, and vision systems are just too much. Very easy to teach.

 

[darrenj]

So there is no gaurunteed break between the bags?
That's tough.
Vision might be your only way to count accurately then.

No guarentee..what does a vision system cost nowaday?

 

[TDMartin]

Base systems, and I'm going off memory, and there are lots of options:
Banner PresencePlus Pro starts at around $2200 US.
Cognex InSight starts at around $4000 US.
DVT Legend starts at around $4000 US.
Cognex Checker (not really a vision system) $1900 US.
They might all include software now.
Some sales guys might get mad at me for being off by a couple hundred or so, but that's a ballpark.

 

[Bob O]

darrenj,

I am thinking aloud here but if you cannot reliably sense the position of the bag, I am not sure a vision system will work for you.

With vision systems, you need correct lighting, lens, fixture and I will add reliable part presentation.

The camera is signaled to take a picture and perform an inspection and in my systems, I use another sensor to trigger the inspections.

If my part presentation were unreliable then my quality inspection would also be unreliable and I have run into this on a shrink labeling line before.



Can you use a through beam across the top of the bag and sense a height difference. I would think that a bag that is tied would be taller then a broken bag and use a timer on delay to eliminate false triggers of the tied end.

I would also think that even if there is not a gap and the bags were running nose to tale there would be a height difference between bags.

If you have tried this or this seems crazy, please ignore.



Example of the cost of vision system from DVT
Legends 510 $2000 [depending on line speed]
Lens $200
Break out Board $200-$300
Off axis light 4x4 white LED $750 or integrated ring light $300
Power cable $50

I do not work for DVT. I am just an end user
Bob

 

[russrmartin]

I will assume that the bags are riding on the conveyor sitting parallel bottom to top with the direction that they are traveling, or perhaps from top to bottom. Either way, I think the best solution is to simply turn the bags 90 degrees as they pass thru the sensing area, then look down from above at the middle of the bag or below. You should easily be able a bag sitting sideways on a conveyor as they pass through. This would also open up the 1/4 inch gap between the bags, as they appear to be taller than they are wide. I will try to illustrate.

----------------------------------
BAG BAG BAG Conveyor Direction ->
-----------------------------------

-------------------------------------------------
B B B Conveyor Sensing Area
A A A Conveyor Direction -> ->
G G G
-------------------------------------------------

 

[russrmartin]

Little Better Visual, but not much

Here's a crude representation of what I think you have, and what I would try.

 

[Terry Woods]

darrenj...

Are the bags end-to-end, or side-to-side?

If side-to-side then, even if they are butted up to each other (but not too tightly), a Banner Laser-Ranging-Eye should be able to easily discern the passing of a bag.

If end-to-end... that might be more problematic.

Also... you might consider a "voting-pair", each from different points of view. If both agree that a bag has passed then you can believe it.

Or, maybe... you are counting in the wrong place... How about counting the bags, up line, as they are filled, or tied?

What is the orientation of the bags?
Can you count in a more appropriate place?

 

[darrenj]

O.K...thanks for the replies..\
First..the bags are end to end..that the problem..
second..after more discussion today we are going to change the eyes to proxies and count the number of times a bag is dropped onto the belt..The "Pouches" are placed into the bag and then the bag is relesed onto the conveyor..This will solve the problem of double counting but wont tell us of bad bags, split bags..etc..they are going to try this and see how it goes..

If this dosnt work as they planned (It will be better but not 100%) i think the light curtain idea would possibly work..

Thanks all

D