Refillable mugs

[GaTechGal]

I think they just need a Marauder's Map.

[LeonardKinsey]

I just heard back from ValidFill and posted about it on my blog, but I'll repost here to keep the conversation going:

Here’s the email I wrote to them:

Hi,

I sent an email to your [email protected] email address, but it was
returned with a “user unknown” 550 error…. Hopefully this address works!

I’ve been looking at your RFID vending products, and had a few questions
that weren’t answered on your site.

1. Is the machine/cup/tag system self-contained, or do we need an external
server? I’m just wondering if all of the data is stored/written to the
RFID tag, or if it writes to some extensive back-end database that we’ll
need to support.

2. If we’re having issues with the system, or need to override for a
customer, how difficult is this to do? Can a cashier do it, or is it
something that only a manager/supervisor could handle?

3. If the RFID tag is damaged somehow, is there an easy process for
restoring the customer’s information to a new tag/cup?

Thanks in advance,
Leonard Kinsey

–

And here’s the response I just received:

Leonard,

Thank you for your message.

I want to apologize for the info@validfill not working. We have corrected that email account and it is now functioning.

Please see my responses to your questions below.

1. Our system is self contained. There is not a requirement for an external
server. All the cup information and data is stored on the RFID tag. There is not a need for a back-end server.

2. There is a bypass switch on the side of our machines to allow a non-rfid tagged cup to dispense. The cashier would insert their key and turn it to the bypass position.

3. There is not a process to restore information to the users cup if the tag is destroyed. We are considering it as a future option.

–

So the main question has been answered here: the data IS written directly to the RFID tag, and does NOT reference a back-end database! As I said previously, this definitely opens up the door for the tags to be hacked, or for “unlimited refill” tags to be programmed and sold by enterprising hackers.

Of course, the easier way to hack the system would be to find a discarded/expired mug and stick it in the microwave, which destroys the RFID tag. Since there is no way to restore the data, they can't prove you didn't buy a refillable mug, so you can complain to Customer Service that your mug isn’t working and have them reissue a new, functional one (you’d have to say that you threw away your original receipt). Stupid guests are always right, right?

Leonard Kinsey
Author, THE DARK SIDE OF DISNEY
The "Anarchist Cookbook" of Disney Travel Guides
http://www.darksideofdisney.com

[ktulu]

1. Our system is self contained. There is not a requirement for an external
server. All the cup information and data is stored on the RFID tag. There is not a need for a back-end server.

I would follow up and ask if you decided to use a back-end server if they support that.

[LeonardKinsey]

I would follow up and ask if you decided to use a back-end server if they support that.

Will do!

Leonard Kinsey
Author, THE DARK SIDE OF DISNEY
The "Anarchist Cookbook" of Disney Travel Guides
http://www.darksideofdisney.com

[ktulu]

Will do!

Leonard Kinsey
Author, THE DARK SIDE OF DISNEY
The "Anarchist Cookbook" of Disney Travel Guides
http://www.darksideofdisney.com

Cool :) My guess is Disney will be spending enough that the company would do whatever they wanted, but this is going to be interesting to see how it unfolds :)

[delsdad]

I just read, on another board, that there are already issues with SGs breaking the activation arms. When the machine wont dispense into an old or expired cup, they just push the arm harder, and SNAP, now no one can get a drink from that flavour.

Also the author claimed it was really playing havoc with the Brazilians, who pass the mugs on from one tour group to another. At least some good can come of this if it reduces the amount of coke those kids consume ! :twisted:

[ktulu]

I just read, on another board, that there are already issues with SGs breaking the activation arms. When the machine wont dispense into an old or expired cup, they just push the arm harder, and SNAP, now no one can get a drink from that flavour.

Also the author claimed it was really playing havoc with the Brazilians, who pass the mugs on from one tour group to another. At least some good can come of this if it reduces the amount of coke those kids consume ! :twisted:

"You just broke our machine. Do you wish to pay with cash, credit, or charge it to your room?"

[Zazu]

If you mean keeping track of how much time you spent where, probably not. It's a different set of problems from tracking customer buying habits and even entries and exits at points of interest. To track the costumes it would take a whole lot of receivers and the amount of data would be huge. To get the right coverage, receivers would have to overlap, and to monitor in any meaningful way they'd have to poll frequently. So you end up with a lot of data points that need to be de-duped (due to overlap) and interpreted. It would be a lot of work and resources for a general thing.

I think this is an slight overstatement of the task. It's not reasonable to track the specific location of every tag at every moment. There are two approaches:

1. Do a phased scan of the property, perhaps every five or ten minutes. It's hard to do anything in that short a window, so all of the locations you visit will get logged.

2. Pre-define the areas of interest, and track only the tags that pass from one area into another.

Right now, this is done with admission media with the "areas" defined as parks. With RFIC, it would be a simple matter of putting sensors in the pathways between each land and capture each tag as it moves land-to-land, much as the E-Pass senses when my car drives from one section of the turnpike to another.

Such "areas" can be arbitrarily small. You could scan each attraction and shop, or even each aisle within a shop. The latter would probably be something done as a temporary survey, it wouldn't be terribly productive to do it daily. (Though Attractions still logs the guest count of every attraction each hour, so perhaps it's easy to underestimate Mickey's appetite for raw data.)

This second approach generates far less data, essentially one data point for each significant movement. De-duping even a 10-minute snapshot of a park would be a pretty huge computational task, and the quality of the data might not be any more useful.

It will be interesting to see how this is implemented and used.

[Alyssa3467]

I was thinking something similar. Just have scanners at doorways and such. A difficulty with that approach would be distinguishing between exit and entry scans, and people loitering at the doorways without actually entering or exiting.

[WEDFan]

Zazu,

We were essentially thinking the same thing, but I didn't express it well. GaTechGal mentioned the Marauder's Map and that was the type of application I was saying probably wouldn't be practical. Real-time tracking of items to within a few sqare yards.

At the end I mentioned a more focused purpose to the tracking being a more likely or possible scenario. This dovetails nicely with the type of restrictions you were writing about. Things like a more limited number of points of interest or point-in-time snapshots. Just the sort of thing I mentioned with tracking customer habits. Short term surveys with a much more granular level.

Neither one of us mentioned that it would also require integrating the costuming system with the tracking system so you could turn an article of clothing into a person. Not a huge task, but depending on what the costume inventory system has available for integration points, could also be pain.

Ultimately, the question becomes whether the effort required would give enough benefit to make it appealing. Does management currently think that there is a signifiant portion of the CMs that aren't where they're supposed to be?