Question W/ New Angle: Soakable Vs Unsoakable US Self-Adhesive Stamps

Hi,

I hope yet another question on this subject will be tolerated. I hope that I have not missed similar threads/answers.

My SPECIFIC PURPOSE is to be able to us NON-philatelic clerical staff members, with extremely limited experience, to sort out which U.S. self-adhesive used stamps on paper are soakable vs non-soakable. The keys to this question is NON-philatelic and extremely limited experience. (The fact that the stamps are self-adhesive has already been sorted out by an experienced philatelist... those that are self-adhesive then need to be sorted into soakable vs non-soakable.)

Before you think about answering this, please imagine yourself trying to train somebody who has never before seen a stamp in their life -- and they may not even be native (or at all) English speakers -- and you have to PAY THEM BY THE MINUTE, thus speed is very important... and if they accidentally put non-soakable stamps in the soakable category, the stamps will be ruined.

Yes, I know that the Scott Catalogue has an icon to indicate which stamps are non-soakable. However, it is not practical, in just a few minutes, to train a non-philatelic person (who may even be an hourly temporary worker) to use the Scott Catalogue. That's just not going to work. This process has to be fast, fast, fast!

The solution to this has to be highly VISUAL, and will also probably include some guidelines to make some of the decisions. It should also require the least amount of work, memorization, or knowledge of English language.

I have been thinking about...

- Making POSTERS that picture the self-adhesive stamps (this sorting is of ONLY stamps that are already sorted out to be self-adhesive) that ARE or ARE NOT soakable, depending upon other guidelines.

- Making simple guidelines based on face value. For example, it seems to me that "most" of the non-soakable stamps started appearing around 2008, in the middle of the 42-cent-era. Thus one could picture on a poster...

... all but these pictured stamps of 44 cents and ABOVE *ARE* soakable. This would define both the pre-2008 and post-2008 stamps that are above 44 cents and happen to be soakable. (Anything else 44 cents and above would thus NOT be soakable.)

... all these pictured stamps of 43 cents and BELOW ARE *NOT* soakable. It seems that majority of self-adhesive stamps from mid-2008 (middle of the 42 cent era) and earlier ARE soakable, thus it is easier to picture what is NOT soakable of 43 cents and below.

... all of these pictured FOREVER stamps *ARE* soakable. Forever stamps started in April 2007 and while some have been soakable, within a couple of years, most have been non-soakable. Thus it is easier to picture the Forever stamps that ARE soakable.

... all of these pictured "non-denominated but other than 'Forever' " stamps *ARE* soakable. In this category are the pre-Forever-era stamps with non-numerical designations such as "First Class", all the various bulk rate, etc., self-adhesive stamps. [I am not positive if there are CURRENTLY more stamps in this category that are soakable VS are not soakable, but as more and more non-soakable "bulk rate" (etc.) stamps are issued, there will eventually be more that are not soakable.]

Other than sheer memorization or using the Scott Catalogue, are other other and/or better ways to do this sorting?

Remember, that the goal is to be able to train, within just a few minutes, to do this sorting, somebody who has never before seen a stamp, and may not even speak the same language you speak.

I look forward your ideas. Thanks!

As a worldwide collector, "sorting" has been my skill, developed over years.

This is how I would handle it.
1. Sort by years
You can have a poster with images of all that years stamps
(eg all year 2000 go in one box)

2. Sort by soakable (again a poster of that years soakables, the rest go into
the unsoakable pile.

Simple really, the crucial point would be identifying either group,
I have no lists for those.

Here is my explorer type hiearchy
I can search either yearly or across multiple folders with one click

If I had a list of soakables, I would add a suffix to the image, such as "y"

I do that with ""surcharges" I can find every surcharged stamps in the US with one click

Same with "NVI"

Same with every 4c stamp etc and so on.

No need to reinvent the wheel.
In the Stamp Smarter database (which most of the US web pages are based upon) there is a field for 'Gum Type'. This field is only shown in some 'detailed' views but 'Gum Type' can be filtered and sorted on in the Advanced View tool.
It would only take me a few minutes to generate a new webpage which allows users to quickly view the self-adhesive issues if desired.
Don

Nice work Don,
Is there a problem with soakable self addhesives, and non soakable
with the US?

I think some Canadian have both.

Thanks for the ideas submitted so far. I appreciate them. However, if I correctly understand those ideas, they require more skills and resources, (and more time to use them) than is available.

Sorting by year requires figuring out what year they were issued. Sometimes easy (with a magnifying glass) and sometimes difficult (and sometimes impossible for a non-philatelist). Also, of course, sometimes on definitive stamps when the year is stated, that is a later year (i.e. later printings) than when Scott considers the original printing of the stamp to have been issued and where Scott has located the catalog number in the catalog. But, regardless, it would be too slow.

Regarding leveraging the Stamp Smarter database, _if_ I correctly understand the concept, the worker would have to know what stamp it is (by number). That is not possible in this case. Using the images in such a database is fine, except that it requires time, but that means "seconds per item, instead of items per second".

Of course I don't expect this work to be done in "items per second", but the only way it can be economically viable is if the work can be done in 4-8 seconds per item (and for hours at a time).

The problem is that the time costs about 25 cents per minute (plus supervision, taxes, etc., etc.) When many thousands of stamps are involved -- and this is only the first stage in a long process -- the cost gets "real" very quickly.

Still hoping for more ideas. That don't involve the worker having to know so much.

Jay

Another idea which has been sent to me off-list is...

OOPS when first I tried to post this I was blocked because of the brand name, thus I have replaced the brand with "****".

Use **** with its SRS module (stamp recognition software) -- and ****'s image separation or SRS's image separation -- to create a large number of individual images.

The SRS can identify the stamp by catalog number (assuming it is in the **** database, which all of these would be) as long as the supplied image format is what SRS is expecting to see.

However, while that does supply (one or more) Scott number(s), that in itself does not automatically supply the most critical piece of information that is being sought: soakability. I don't know if ****'s database is pre-filled with that data for modern U.S. stamps. (I will ask.) I also don't know if the SRS process can be tweaked to DISPLAY, as part of the same process, that specific bit of information (soakable yes / no).

Of course this approach still presents a couple challenges:

- The time to prepare acceptable images for the SRS process.

- The image separation process (to make individual images for SRS to identify) is expecting the stamps to be on a black background. However, these stamps are still on paper and the paper they are on is not black. Thus the images may have to be further manipulated to get something that can be processed in that workflow.

- The stamps CAN be manually, individually cropped out to individual images (the perforations are not very important for the SRS process), however, individual cropping-out is time consuming work and does require some skill and very high speed of working on a computer that allows for fast "mouse pointer" manipulation (drawing the boxes around stamps for manual cropping) -- probably not something that can be done at high speed on a phone or tablet.

What is intriguing about this concept is that:

a) It suggests that with further development, much of the work could be done remotely -- a big benefit in finding people to do the work (inexpensively).

b) Again, with further development and proper workflow design, it could possibly be almost completely automated once a page-full is scanned, eliminating the need for people. However, this would require further software development to "report out" the "soakable yes/no" data to match up to the PHYSICAL scanned page of stamps -- somebody then has to take the stamps off that page and put them in physical piles of "soakable yes/no". Because this is not all "just data" -- the physical stamps have to be moved to the correct pile -- it is more complicated than it sounds to automate the ID process and then connect the resulting information with each of thousands of physical stamps. I am sure it can be done -- but I am not so sure that the software could be developed economically enough.

With respect, you're missing the point.




You don't need any of that
Visual image recognition is the easiest and cheapest route

You make a poster on the wall with every stamp of a given year
with a box underneath to accept the stamps

say 1978 or 2015 or 2020 or what ever
Make the poster 3 feet square or whatever
1978 comms

Sounds like a job for Mechanical Turk.

There is a member here on SCF, that buys huge Kilogram lots of
Australian, and sorts them into piles of like stamps and values
The family does it, from what I gather,
it's not that difficult
A complete newbie, after day of sorting, should be able to manage
400 or 500 an hour

Hi Rod222... You Said:



Let me be sure I understand... You propose that there should be approximately 20 posters on the wall... and that the posters show ALL the self-adhesive stamps ... and NOT use logic to eliminate those which can be filtered (i.e. answer known) just based on denomination and a more limited group of images to check for that denomination?

Also, I don't know what you mean by "... with a box underneath to accept the stamps". We are not trying to "identify" the stamps or sort them by years, etc. We are trying to put ALL the stamps into ONE OF TWO BOXES: Soakable: YES vs NO.

That method would seem to have to require the worker to know what year the stamp was issued -- or check multiple posters to find which year it is in.

Why is that faster?

If ALL self-adhesive stamps are going to be on the posters, then it would seem to be faster to do them by denomination rather than year.

Maybe I am missing something more fundamental about what you are proposing. If so, please advise.

[I should add that since these are still on paper, that some pieces will have multiple self-adhesive (but ONLY self-adhesive) stamps on them. That means that the worker will have to cut up some pieces to separate soakable from non-soakable if both are on one piece.]

Beginning to sound like the best course of action is to soak them all and forget sorting.

Then I would keep it as simple as possible.
Water-activated-gum vs self-adhesive. Period. Simple.
Do not fret about identifying the small number of soakable self-adhesives.

(Frankly, I wouldn't do this for minimum wage for hours on end.)

What is the ultimate end of this process?

Rogdcam said:



Thank you very much for this. I was not aware of that particular service -- I had to Google it. Very cool and there is also some other work for which I may be able to use such a service. That is going to take some exploration.

Okay, apparently I did not successfully make my goals clear and the limitations clear. Referring back to my original post and later posts:

GOAL: Separate U.S. self-adhesive stamps into two groups: soakable VS non-soakable.

LIMITATIONS / REQUIREMENTS:

a) Must be done at extremely low cost.

b) Must be done for many thousands of stamps. All the stamps have already been confirmed to be self-adhesives.

c) Workers doing the work know nothing about stamps; may never have seen a stamp before; and may or may not speak your language. Their training cannot take longer than a few minutes. (Thus a reliance on images and logic for what is and what is not soakable.)

d) Stamps are on paper; each piece of paper could bear more than one self-adhesive stamp, thus the worker may have to cut apart some pieces of stamps are in more than one category.

e) Stamps that are non-soakable must NOT be soaked -- soaking a non-soakable stamp will usually destroy it (in terms of collectible quality). [Somebody mentioned just try soaking them all; that will destroy those that are not soakable. Somebody may have mentioned don't soak any of them; the soakable ones are only salable if they are soaked.]

f) The number of U.S. self-adhesives that ARE soakable is NOT small. Prior to 2008 many/most self-adhesives were soakable. During 2008 it is about half and half. After 2008 most (but not all) are non-soakable.

g) This subject has nothing at all to do with traditional NON-self-adhesive stamps. The items that need sorting into soakable vs non-soakable have already been sorted to only include self-adhesives.

Jay - I think we all understand what you need to accomplish. The driver of the enterprise is labor cost and only you know what production/pay rates you need in order to turn a profit or at least not lose money. Perhaps those rates need to be reexamined? It sounds as if you backed into the desired end result rather than starting with what you can practically accomplish and seeing where you end up. Sometimes the square just does not fit in the circle.

Why is that faster?

Well, I was assuming that you were receiving packages of Kiloware.
So, I see you are already working with self adhesives only.

So how many years would that cover?

Sorting by years first, it isolates and discriminates on
the piles of self adhesives.

Your going to need a large table, spread the stamps out and
place in piles from visual similarity.

Then place in YEAR boxes, then sortable by soakable and not.

Faster?
because the end result, if I am correct, you'll want finished stamps
in year and catalogue number, this system has that in its route.

Otherwise you have a pile of soaked and non soakable which you will
have to sort again anyway. (double handling)

Unless your handling in the hundreds of thousands, a simple pile
of say 1000 to 2000 or even 5000 should be easily managed manually.

When I receive a swap of 100 to 500 stamps worldwide,
I discriminate by COUNTRY
That country by value tablet
That value by year
Then to album.

I have never found a faster way

My imagination sees a part time worker probably needed for
1 day a week, or fortnight to accomplish enough stock to
keep you busy with all the facets of bringing them to market.