Building a Bitcoin ATM. Should I bother? Demo Video!

[JayCoin]

I have been building a bitcoin automatic teller machine which consists of a computer, webcam/QRcode scanner, bill acceptor, touchscreen and receipt printer.  The idea of it is to allow anyone to purchase bitcoins with cash just by scanning their bitcoin address and depositing cash into the machine.  The machine then instantly sends bitcoins to the address at the current exchange rate plus some minus a fee of some sort.  Has anyone done this yet? Does anyone think that this is something people will use?

I will eventually post a video showing it in action once I fix a few more bugs in the user interface.  

EDIT: Here is a crude demo video of the prototype http://youtu.be/EdR_fHr_6Eg    

[casascius]

How reproducible will it be?  It's cool if you're able to produce a one-off machine to show at tradeshows, but if you can produce something that others can build just by buying the same components, then your efforts will be that much more fruitful.

I had thought of a design that does much the same thing: it's what you've described minus the bill acceptor: https://en.bitcoin.it/wiki/Casascius_Bitcoin_POS_system . It replaces the bill acceptor with a human operator, assuming the purpose of the device is to accept bitcoins in exchange for goods and services, and optionally to dispense them in exchange for cash.

It's vaporware, but it's also a significant enough idea that the European Central Bank thought to include a link to it in their recent virtual currency whitepaper.

[repentance]

I have been building a bitcoin automatic teller machine which consists of a computer, webcam/QRcode scanner, bill acceptor, touchscreen and receipt printer.  The idea of it is to allow anyone to purchase bitcoins with cash just by scanning their bitcoin address and depositing cash into the machine.  The machine then instantly sends bitcoins to the address at the current exchange rate plus some minus a fee of some sort.  Has anyone done this yet? Does anyone think that this is something people will use?

I will eventually post a video showing it in action once I fix a few more bugs in the user interface.      

Where are you planning on locating your ATM that it will get enough traffic to justify the overheads of operating it?  How are you going to keep the machine secure from vandalism/theft?  How are you going to calculate your hot wallet needs so the machine isn't constantly going offline because your hot wallet has no more BTC - unlike a bank, you can't just go to another ATM. The idea has been proposed before and there are a number of obstacles which need to be considered in order to make it viable.

[casascius]

I think a reasonably viable (not to mention exciting) proof of concept would be an acceptor module for a snack vending machine that scanned QR code private key, allowed a snack purchase, and then returned the change back to the original bitcoin address.  Only the payment acceptor would have to be modified in any way to use Bitcoin - the remainder of the entire machine would be a standard unmodified vending machine of the sort already commercially available.

This is exciting to me because it's practical, has a clear useful purpose, and is reasonably possible with off-the-shelf hardware.  The scope of the invention would be limited to just the payment acceptor.

[casascius]

Here is another way to leverage the work of others:

Look at this prepaid phone card vending machine: http://www.onlinevending.com/catindex/items/C100-19-0.html

Simply load it with empty paper wallets with a Bitcoin address on them, pre-printed onto standard size plastic cards.

All it requires is that you have some way to know immediately when a card is sold (for example, a sensor that notifies your server of a sale via a cellular connection).  Your server automatically calculates the correct amount of BTC and sends it to the card at the time it is vended.  Assumption is you know in advance the sequence of the cards so you know which one just got sold.  Voila, bitcoin ATM with minimum engineering.

With even less engineering, you could activate the cards simply by pre-printing activation codes and then having the buyer use their cell phone to notify you they just bought the card by visiting a URL, scanning a QR code, or calling an activation number just like a prepaid phone card.  (they end up getting the exchange rate as of the time they activate it instead of the time they buy it, but it's sixes)

[JayCoin]

Where are you planning on locating your ATM that it will get enough traffic to justify the overheads of operating it?  How are you going to keep the machine secure from vandalism/theft?  How are you going to calculate your hot wallet needs so the machine isn't constantly going offline because your hot wallet has no more BTC - unlike a bank, you can't just go to another ATM. The idea has been proposed before and there are a number of obstacles which need to be considered in order to make it viable.

Where are you planning on locating your ATM that it will get enough traffic to justify the overheads of operating it?
I was hoping to find some people from the forum that own businesses that would be willing to place a machine in their business.  They would, in return, get a portion of the fees.

How are you going to keep the machine secure from vandalism/theft?
Same issue as any vending/atm machine.

How are you going to calculate your hot wallet needs so the machine isn't constantly going offline because your hot wallet has no more BTC - unlike a bank, you can't just go to another ATM.
The machine doesn't hold the wallet.  All the transactions take place on a remote server. (Machine will need internet connection). Remote server would monitor wallets and make sure enough bitcoins are available.

I have searched the forum for other discussion about this, but I can't seem to find any posts.

[casascius]

Here's a couple examples of places where Bitcoin "ATMs" (or prepaid card dispensers) would work well for starters:

1. Internet cafes
2. Tradeshows with lots of tech people in attendance
3. I know it's not kosher to mention places where Silk Road customers would frequent, but got to admit that the machines would get used by that crowd wherever they can get to them.

Suppose I set up shop and wanted to get into this business... I wouldn't even have to bother with machines.  Merchants can buy their own machines or vend the cards directly from the register by hand.  I simply sell prepaid cards to merchants (internet cafes) that can be vended for $10 each... each card is essentially $9 (or some similar value) worth of "Casascius Dollars" (think MtGoxUSD) where the buyer can instantly convert them to BTC and withdraw them to the address of his choice via cell phone.  By selling "Casascius Dollars" the merchant doesn't have to think of them as anything other than dollar-denominated prepaid cards nor worry about exchange rates since the actual exchange into BTC goes into effect the moment the buyer cashes out the card.  The remaining $1 or whatever is the merchant's profit.

I doubt I'd have an interest in making such a thing but this would be a perfect thing for BitPay to do...

[MPOE-PR]

Well to answer your original question: whether your efforts with this come to anything or not doesn't hinge on whether the idea has merit (it does), but on whether you have the resources and drive to make it happen. It's one of those things that are 1% inspiration and 99% blood sweat and tears I think.

[Isokivi]

I have been building a bitcoin automatic teller machine which consists of a computer, webcam/QRcode scanner, bill acceptor, touchscreen and receipt printer.  The idea of it is to allow anyone to purchase bitcoins with cash just by scanning their bitcoin address and depositing cash into the machine.  The machine then instantly sends bitcoins to the address at the current exchange rate plus some minus a fee of some sort.  Has anyone done this yet? Does anyone think that this is something people will use?

I will eventually post a video showing it in action once I fix a few more bugs in the user interface.     

Consider seeing this device at an airport as you are about to leave the country, with a random amount of local currency in your pocket that you'd have to exchange for your local currency at some point.

[tehace]

Maybe a much more convenient design would be to have some sort of vending machine with paper wallets in some kind of  xray, etc protective shielding. The paper wallets could have common amounts of bitcoins eg 10 btc, 1 btc 0.1btc and the user could simply purchase them the same as they would purchase a bag of chips. This eliminates the problem of a hot wallet, simplifies the process greatly on your end and  provides instant feedback on a purchase. It does however introduce the problem that you are storing hundreds if not thousands of dollars worth of bitcoins to start with and fiat to end with  in a vending machine.

[casascius]

Maybe a much more convenient design would be to have some sort of vending machine with paper wallets in some kind of  xray, etc protective shielding. The paper wallets could have common amounts of bitcoins eg 10 btc, 1 btc 0.1btc and the user could simply purchase them the same as they would purchase a bag of chips. This eliminates the problem of a hot wallet, simplifies the process greatly on your end and  provides instant feedback on a purchase. It does however introduce the problem that you are storing hundreds if not thousands of dollars worth of bitcoins to start with and fiat to end with  in a vending machine.

This would work very well other than it adds the problem of the machine having to keep track of the exchange rate, as well as accept odd amounts (e.g. instead of inserting a $10 bill and getting 0.81 BTC, you would instead be inserting $12.27 to get 1 BTC, and would need a coin acceptor and change dispenser, adding significantly to the equipment and maintenance costs).  Not insurmountable by any means, it's just more cost and complexity... if I were building it, I'd find it easier to build a website that funds cards when a code is entered (even with the risk of a hotwallet), vs building some sort of realtime communication into the machine so it always knew what to charge.  On the other hand you can order the phone card dispenser TODAY and simply cover over the artwork with new artwork.  If I ran a service that supported thousands of these machines, I'd still only have to maintain one activation server and one hotwallet.

BitPay...this is the perfect product for you to sell.  Prepaid cards with "BitPayUSD" on them, where someone can go to a URL and have the current value in BTC sent to the address of their choice.  If cards were available for sale, I'd probably buy the vending machine just as a novelty and to take to places where bitcoin buyers might be interested in buying BTC.

[tehace]

Maybe a much more convenient design would be to have some sort of vending machine with paper wallets in some kind of  xray, etc protective shielding. The paper wallets could have common amounts of bitcoins eg 10 btc, 1 btc 0.1btc and the user could simply purchase them the same as they would purchase a bag of chips. This eliminates the problem of a hot wallet, simplifies the process greatly on your end and  provides instant feedback on a purchase. It does however introduce the problem that you are storing hundreds if not thousands of dollars worth of bitcoins to start with and fiat to end with  in a vending machine.

This would work very well other than it adds the problem of the machine having to keep track of the exchange rate, as well as accept odd amounts (e.g. instead of inserting a $10 bill and getting 0.81 BTC, you would instead be inserting $12.27 to get 1 BTC, and would need a coin acceptor and change dispenser, adding significantly to the equipment and maintenance costs). 

I was thinking more along the lines of a fixed exchange rate that charges a premium above spot with set prices ie a 1btc card for $15 the machine is competing against local bitcoin traders who can sometimes charge way over spot for the anonymity and convenience. I see no reason why an even more convenient option couldn't charge an even greater premium. Im just trying to come up with an absolute minimum setup thing here.


[/quote]
BitPay...this is the perfect product for you to sell.  Prepaid cards with "BitPayUSD" on them, where someone can go to a URL and have the current value in BTC sent to the address of their choice.  If cards were available for sale, I'd probably buy the vending machine just as a novelty and to take to places where bitcoin buyers might be interested in buying BTC.

[/quote]



Quoted for truth.

[deepceleron]

How about a business like CoinStar machines giving out BitCoins. They already have a business model charging dummies 9.8% for exchanging USD->USD. You'd only need a better way of giving machines your address than typing in 34 digits, a "print qr code for this address" in Bitcoin is a first step. Make a good usage model where interacting with the machines isn't hard (starting with a universal "pay to me" card that can give differing machines your bitcoin address) and they might consider it.

"An estimated 84 percent of coinstar users are women". Not going to comment on this one.

They recently sold off their money transfer business to Sigue, a company that profits from what Bitcoin does for free (http://www.sigue.com/index.php?option=com_content&view=article&id=122&Itemid=114), so they don't have a conflict of interest in now promoting an instant and free money transfer method.

[Akka]

There is already this: http://blog.maschinenraum.tk/2012/07/15/bitcoin-vending-machine-exchange-euro-coins-for-bitcoin-wallets/

it gives you a paper wallet for cash.

Was done by this user: https://bitcointalk.org/index.php?action=profile;u=62162

Hasn't been online since then, but maybe his PM mail works?

[420]

what premium will u charge

[JayCoin]

I haven't figured the economics quite out yet.  I think I would charge very little in the beginning to develope a customer base.

Here is my prototype demo video http://youtu.be/EdR_fHr_6Eg

[n8rwJeTt8TrrLKPa55eU]

I haven't figured the economics quite out yet.  I think I would charge very little in the beginning to develope a customer base.

Here is my prototype demo video http://youtu.be/EdR_fHr_6Eg

Fantastic video!  Much more exciting than any written description.

You could probably put it on Kickstarter and get funding.

Will be twice as impressive if you can ever demo two-way conversion between BTC & USD.

[Remember remember the 5th of November]

I haven't figured the economics quite out yet.  I think I would charge very little in the beginning to develope a customer base.

Here is my prototype demo video http://youtu.be/EdR_fHr_6Eg

Fantastic video!  Much more exciting than any written description.

You could probably put it on Kickstarter and get funding.

Will be twice as impressive if you can ever demo two-way conversion between BTC & USD.

The only problem that arises is how the buyer will know he will actually get the coins and the money won't be sent elsewhere.

[JayCoin]

I haven't figured the economics quite out yet.  I think I would charge very little in the beginning to develope a customer base.

Here is my prototype demo video http://youtu.be/EdR_fHr_6Eg

Fantastic video!  Much more exciting than any written description.

You could probably put it on Kickstarter and get funding.

Will be twice as impressive if you can ever demo two-way conversion between BTC & USD.

I maybe able to do small amounts the other way with zero confirmations (No one would want to stand infront of an ATM for 10 minutes waiting for transaction confirmations) , but the fees would probably need to be high to cover any double spend fraud.

[SgtSpike]

I like it!

Another idea is, for those who own or manage vending machines, stock them up with 0.1 BTC paper wallets and charge $2.00 each.  If the price of BTC changes to > $20, then break them down into 0.05 BTC paper wallets, etc.

I would so do this if I owned a vending machine.  Heck, I'm thinking of buying a vending machine just to do it.  Lots of them for sale (even active ones) on craigslist, etc.