NYT: Wal-Mart Slows RFID Plans, Suppliers Resist

securitas writes "The New York Times' Barnaby Feder reports that Wal-Mart has scaled back its plans to deploy RFID tags because the majority of its top 100 suppliers will not be able to meet the Jan. 1, 2005 deadline that the retailer demanded. Suppliers are resisting Wal-Mart's RFID demand for a variety of reasons according to AMR Research. Only 40 suppliers will meet the deadline, with two suppliers 'so tied up in a complete overhaul of their entire information technology infrastructure that they have put off attempting to introduce radio tagging.' A more pragmatic reason for the delay is that 'no one who uses the technology has systems that can reliably read the information 100 percent of the time in factories, warehouses and stores; Wal-Mart said the rate was around 60 percent in its stores.' It's hard to make the case that RFID will help track inventory when you can't reliably find 40% of it."

so who benefits more

[Anonymous Coward]

as a % out of this situation in profit and capital

Customer
Wallmart
Distributer
Manufacturer

then perhaps you can understand the remaining parties reluctance to make the expenditure

Roles reversed

[asliarun]

Given that Wal-Mart has been bullying its suppliers since donkey's years, it's high time they got a taste of their own medicine. However, rumour has it that the Pedigree has pawed the line in this initiative. Only, they're calling it Arf-ID.

cough, sorry

Re:Wait...

[clone22]

Suppliers were due to have ability to tag at the pallet/case level. At current tag prices that's not to bad of a hit. However, the tags are the trivial part of an implementation. It's getting the data from the tags into your information infrastructure that will kill you. I've seen demos where pallets of case goods were trucked S-L-O-W-L-Y through an array of antennas and even at that rate one or more items on the manifest were not recognized. So, what do you do? 1) Check off each item visually? 2) Run around the warehouse looking for the missing items? or 3) Read the boxes with a barcode scanner? Bentonville, we have a problem.

RFID is not cost effective and is very problematic

[lewscroo]

I am currently working with one of the RFID companies that is "working" with walmart on the actual implementation of RFID. Let me tell you that there is no foreseable ROI in the near future. Currently at a cost of about 25 cents a tag, it is much too expensive to be worth it for anyone. The technology is in its infancy so there are so many problems we have encountered so far.

One of the problems is the tags. Not only do they cost so damn much, but they are also not very high quality. There's a feature called "locking" which allows you to set a number on the tag and not allow it to change, but when using this we have too high a failure rate to be effective (10-30% depending on the tag type). So we had to turn off the locking, meaning its much easier to change the unique number associated with the tags (which will be a problem when tags hit the retail sector) and now we only get around a 1-2% failure rate. But when doing high volumes, even this small percent is expensive to deal with.

Another is the hardware. Part of the tag writing problems we have seen may be due to the tags and/or the reader/writer units. But right now, some tags get created and written to with no problems, but when they go by a reader, the reader just does not see a number on that tag, meaning as i said before its either a bad tag or some sort of incompatibility/problem with the reader unit. Currently we are trying to get the tags applied cost effectively, but unfortunately its pretty much boiling down to using people to grab tags from a RFID printer and hand-apply everything.

We have also been having trouble verifying all the product on a pallet, and certainly cannot expect to read 100% of product 100% of time. Some product is easy to see, but depending on the density/material in the materials on the pallet, it can be very difficult to read many of the tags.

Software is another hinderance. While the company i have been working with has had its large share of problems in the last few months, they are getting better, but still are not perfect. And unless things work perfect, it can cause so many problems. One small chink in the software can make it inoperable (essentially crashing the software a-la Windows), but the software is slowly getting more and more stable.

The fact that Walmart madated this is certainly causing issues, especially for smaller companies and products that companies make almost no money on anyway. For us, we have a very expensive product so tagging at the case level is not too big a deal (it still has/will cost us millions of dollars to do), but just remember theres lots of companies that make almost no profit on the case level and that 25 cents for a tag eats pretty much all of their profits. RFID isn't going away, theres just too much potential. RFID can certainly work as a technology, as seen in the success of toll-tags like EZ-Pass and Smart-Tag. And many of these problems would have arisen anyway in the future, its just that the Walmart mandate basically caused the problems to happen faster.

Re:Wait...

[djupedal]

I just went over costs with a local RFID vendor, and the costs are not as cheap as I was led to believe. Couple this with proxity challenged systems, and it's another solution looking for a problem the way I see it. Maybe later, but not now, thanks.

Re:Am I the only one who likes RFID?

[swillden]

There are a lot of reasons for consumers to not want RFID tags.

Perhaps, but you can't really think effectively about this with as much confusion about the technologies as you have.

Would you want to literally broadcast the fact that you had thirteen platinum Visa cards in your wallet?

1. The chips being put in credit cards are not RFIDs, they're contactless smart cards. There are many technological differences but the main practical difference is range. Contactless smart cards have a practical range of about four inches. Actually, I've rarely been able to read one at that range. Typically, to get a reliable read you need to get the card within about 1 cm of the reader. In practice, the most convenient thing to do is to lay the card on top of the reader.
2. Contactless smart cards and readers can't operate with multiple cards in range of the reader. If you put a stack of them on a reader sometimes the reader will be able to talk to one of them but usually none of them will be readable.
3. Contactless EMV cards don't provide any information about the type of card they are, so there's nothing to distinguish between a starter card with a $500 credit limit and a platinum card.
4. If the banks are smart and go to contactless cards with a PIN, the card will refuse to divulge any significant information until the PIN has been transmitted to it. PIN transmission is only done in a secure channel (encrypted).

So, your hypothetical barroom scanner must:

1. Get his reader within a foot (let's be generous) of your wallet.
2. Hope you only have one card in there, because otherwise odds are good he'll get nothing.
3. Have the keys necessary to establish an encrypted session with your card.
4. (Maybe, depending on configuration) Get you to tell him what your PIN is so he can send it to unlock the card.
5. Grab your card number and then use his hacked access to that bank's computer to find out whether or not it has a high credit limit.

I think he'd be better off looking at your clothes and car to see how much money you have.

There are also other privacy related reasons you might not want RFID tags in your clothing. What if you walked into a fancy restaurant and they scanned you on the way in, realized you had on Walmart underwear, and refused to serve you?

Well, retailers who plan to use RFID (and these are RFID tags, not smart cards) also plan to deactivate the chips at the checkout stand. Among other things, that will allow them to identify items that have been stolen, rather than purchased.

Or, would you want that restaurant to throw you out before they seated you because they saw your Visa cards were maxed out?

In this case the restaurant would have to do everything the barroom scanner would, plus perform a credit check. EMV cards don't provide (don't really even know) your credit balance. I suspect the restaurant's jet-setting clientele would get irritated at all of the extraneous credit queries. What would the restaurant do when someone walks in who has their credit records blocked?

There may be privacy concerns with RFIDs and contactless smart cards, but your examples are both infeasible and, frankly, rather silly.