Machine that sorts coins by date

I'm a software engineer by trade, and I am considering building a machine which can sort coins by date / mint. What I am trying to figure out is if it will be worthwhile for me to spend time doing this, from a financial perspective. I would expect the machine to be able to process several coins per second, and I would need to find a source for coins; maybe a deal with Brinks? Or I could sell the machines. What do you guys think?

 

Log in or Sign up to hide this ad.

That would be a cool invention! I wouldn't sell it if I were you. I'd keep it a pull all the good coins out of circulation and sell them for less than a dealer. That way, you're happy AND the buyer is happy. I know you want to gift one to me, by the way...

 

Interesting concept, guessing you'd just set up a database with known coin images and have it compare? If so you could theoretically have it look for errors/etc. Not sure how you planned on having it work but it'd be cool to see

 

Yes, ramedog, you are right. The system will need to have a database of images to compare with - two for each style variation (front and back).

 

I don't believe it would be a wise financially, as the only coin that some collectors want separated are pre 82 coppers for their "melt value".

For this to be feasible you would need a high speed camera,which h would put the price out of the reasonable realm of most collectors. Banks or brinks I doubt care if their rolls are solid date/mm.

 

Yes, you are right KoinJester; It would be an expensive machine - definitely in the thousands.

 

You would have to have much more than two images per side. A few dozen, at least, and enough coding for the machine to tell if something that does not look like any of the images is a dated coin.

 

wkw427, I know how to make the machine, but I am trying to figure out if it is worthwhile. You would need an image of each side, with the date and mint marks removed from the image. You'd rotate these images into multiple images, in memory; maybe an image for each angle between 1-360. You'd use edge detection, and match the closest image. If the closest image is not within a certain tolerance, you put it into the reject bin. At this point you know where the date and mint marks should be, so you can do OCR to determine the date / mint mark.

 

My answer's the same, but it depends on what you mean by "financial perspective". If you mean building a machine for personal use, probably not. It would more than likely take years to find enough good coin to pay for the machine. Factor in the cost of transporting your rolls to search. I doubt Brinks or any other entity will sell you coins direct. And you'd have to return all the coins that don't pass muster.

If it's selling the machine, think about the market. Who'd buy it? It would again, take years for a machine like that to pay off for the buyer.

Not trying to discourage you from building this, but look it as a hobby and not as a profit-making venture (kind of like us woodworkers). I could be wrong, but I just don't see a lot of profit in the venture.

 

FWIW, I'd be willing to spend up to $5,000 for such a machine. On a similar note, I'd be willing to spend EASILY $10,000 for a program that automatically surfs eBay Morgan dollars and analyzes the photos to detect valuable VAMs.

 

What if you are using gem type coins? It then would not match with a heavly worn coin. Likewise if you base it off of a worn coin; it will not pick up an UC coin.

you would have to have at least 360 images, per side, at varying states of wear

 

There is only one way to find out..........Build that puppy

 

I'll take 2

 

You are not the first to think of this, but I hope you are the first to make it so

 

Once the word got out on the "world wide web" i am sure you would have takers..I hope you do this and i am very interested to see a working unit.!!

 

Maybe he should start taking pre-orders now, and use that money to fund the venture. It's easy enough to say "I'll pay $xxxx" or "I'll take a couple" on a web forum, and say once word gets out that it'll take off. Lot of time, effort, and capital involved for the OP to gamble with. Again, if he can afford it, and has the know-how, by all means build a prototype. Keep in mind there are too many variables to consider when compiling the database and writing the code. Not to mention attorney and patent fees (which can add significant cost to the project). Don't forget a little something for the OP in the price (you know, profit ). Will it work? I'm sure the technology exists. Can it be made affordable to the average person, and will that price offer a reasonable return on that investment? I have my doubts.

 

I am sure there would be many applications and I also would like to see it. I believe you could destroy the coin roll searching market in less than a few months if u did it right. You could make the machines so they were inexpensive and easily scalable, but the software would be the key piece and only you would own that. You could sell rights to certain parts of the country or states, then you could sell the licenses to the software and that's where you'd make like a Bajillion dollars. So for example I will buy 4 machines please, 4 territories (ill take NY, Atlanta, Chicago and Los Angeles). Then I'll buy a license to support 4 machines please. I sent the deposit to your PayPal Gift for $4 Bajillion dollars. Send me the contracts and I'll have my lawyer take a look.

 

Cool idea. The only problem I see is the quantity of bins, tubes?, it would require once the coins were sorted. Several hundred.

Maybe a machine that you put just a handful in at a time of the same denomination.
The dates would show on the monitor and you could decide if you wanted to take a look at just that small batch. Just a thought.

 

The way I am thinking of doing it, is to allow the operator to specify which date / mint combos they are looking for, in the sorting program that controls the machine. Then there will only be 3 bins: Accept, Reject, and Unknown.

 

I will wait until I have a working prototype before taking pre-orders.