2015.02 | RFID self-checkout

IMG_8720For many years, retailers have heard about the benefits of RFID, and there has been little to no use of item level RFID to check out in a store.  On the weekend, I visited my local library with my family and had an opportunity to utilize the newly installed self-checkouts to check out our books.

Under the old system, all items had a unique barcode and an EAS security tag.  To check out an item, customers presented their books and library card to an attendant at a PC with a scanner.  The attendant scanned the library card to validate the customer’s identification, checked if there were any fines, holds etc, and then scanned all of the items to be checked out.  A receipt was printed and the
Photo 2015-02-09, 7 16 14 PM items were walked around a security gate at the combined entrance/exit, and handed to the library customer.  The security tags set off an alarm when they are near the gates, so the attendant passed them to the holder on the other side of the security gate.

Under the new system, there is no need for customers to interact with an attendant unless they have a fine, a hold, or some other intervention that goes beyond the simple checking out of an item.  The customer uses one of a few self-checkout terminals – which include PC and a scanner, but also a customer facing touchscreen, and an antenna in a pad on the counter.  The Photo 2015-02-09, 7 16 06 PMcustomer scans their library card, and if they have no fines or holds, they can then identify the items to check out.  To identify the items for borrowing, the customer places the items on the pad on the counter, with as many as three at a time piled on top of each other in a stack.  The antenna reads the tags in the books and shows them on the screen for verification.  The customer can validate that the items match, ensure any media inside the item matches the case, and complete the checkout with or without a receipt.

As the items are now considered checked out, the customer can walk past a gate with an RFID reader, and if the item they are carrying is checked out, no alarm will sound.  If an item has not been checked out and allocated to the customer’s card, an alarm will sound.

On the whole, the system worked very smoothly. While only recently installed, customers took to it and had little issue using it.  The library staff were helpful and encouraging for the few customers who did require assistance.

The RFID system was a good fit in the library for a number of reasons:

  • Even if RFID tags are more expensive than barcode labels and EAS tags, items are tagged once and then run through the system many times as they are loaned through their useful life, instead of being purchased once.
  • Most items in a library are flat and lend themselves to easy scanning on a pad like this.  There are rarely very heavy, bulky or oddly shaped items to be checked out.
  • Consumers are accustomed to self service from using it at ATMs, airports and retail self-checkouts so they recognize the paradigm of self service took to it well.
  • All customers are identified with an identity card.  No exceptions.
  • One attendant can now support many customers at a time instead of one, reducing wait times, and ideally enhancing the customer experience. If customers don’t wish to check out their own items, they can assist them easily.
  • Allowing customers to walk out the door with their own items without passing them around a security gate appears to provide better flow and a smoother checkout experience.  It also removes an underlying assumption of distrust implied by the gates and security system as previously configured.

The whole system reminded me of a question I had from a retailer at NRF who asked to see our RFID self-checkouts.  While I would personally like to see RFID checkouts in retail purely from a love of technology, it seems unlikely at present.  There are some differences between this library scenario and many retail environments from a checkout perspective:

  • It will be difficult to convince all the parties involved in manufacturing goods to move to item level RFID tags unless the retailer and consumer are willing to absorb the price.  The prices are getting lower and lower for the sort of passive tags needed for items purchased at retail.  Time will tell if it will be enough!  The big retailers will have to drive this adoption.
  • Implementing readers to read these tags instead of barcodes would require a replacement or at least an upgrade to current reader infrastructure.  An ROI is needed to change/add that infrastructure to include RFID readers.
  • While libraries have basic flat items, other retail environments have all sorts of uniquely sized and shaped items that may not lend themselves well to a standardized rfid reader environment.  For unusual items, a handheld RFID reader could be used, but if it was, what’s the difference between holding that RFID reader to a tag and scanning a barcode as is done today?  Not much.
  • Would there be a throughput advantage?  For smaller transactions, it is very unlikely.  Cashiers and even customers scanning themselves can scan a few items relatively quickly.  For smaller transactions, tendering is generally the longest part of the transaction and not the scanning.  For larger transactions, there may be some throughput advantage, but it would take time for retailers and consumers to develop the trust that the system would capture all of the items accurately.  Also, many customers like to validate their purchases and their prices as they are scanned.  Much of the throughput advantage of an instantaneous cart total could be lost by questions and validation afterward.
  • Weighable items would need to be either pre-packaged or separated for checkout.  Weighables couldn’t just be left in the cart for reading.
  • Unlike a library, the items that a consumer buys can’t be “checked out” as they are not unique in the store.  An antenna at the front of the store placed for security can’t identify items as valid to pass or not.  It’s unclear how security would work with RFID tags.  I’ve heard of more sophisticated and more costly RFID tags that can be de-activated on scan, but then what if a client changes their mind after the transaction or returns an item?  Does it need to be retagged?  How does one “print” a new tag for an item?  If RFID can’t be used for security, then EAS would be needed as well.
  • What about bulk items that are tagged with paper tags today?  What about low value and small items like greeting cards?

As with all solutions in a retail environment, there must be a benefit for both the consumer and the retailer for a solution to be implemented successfully.  It’s possible that the RFID self-checkout could get to that point if retailers can leverage the operational benefits on the back end first and push it to the front end.  Then it will take customers and retailers getting comfortable with wheeling a big basket of groceries up to a reader and taking that price as correct.

A lot of stars have to align for an RFID self-checkout to come our way, and if they do it will probably take some time.  Maybe next NRF.

Advertisements

2010.13 | Barcode of the Future?

There was a lot of press over the last week or so around RFID tags that can be printed onto packaging  using Carbon Nanotubes in the ink.  This is an interesting development in the world of scanning in retail, given our ongoing obsession with attaining the nirvana of checkout – the non-checkout – as depicted in the famous IBM “You forgot your receipt” commercial from the late 90s.

A few thoughts to consider on using item level RFID tags at point of sale:

Labelling Complexities – Considering the thousands of items for sale in today’s retail stores, expect a migration to RFID to take some time to complete.  For example, while marketers may think RFID means the end of unsightly barcodes, it will be necessary to have some sort of visual code or number on any item in order to allow for pricing in the event of a failed code or reader.  Not having a visual code to type in would cause bottlenecks at point of sale and customer frustration.   Another concern is how RFID readers will recognize items vs cases of items.   Cases of Coke generally have individually barcoded cans as well as a barcode on the case.  Barcodes are visual and the scanner can be directed at the code to ensure it picks up the case code.  Should Coke not put RFID tags on cans in cases?  How do we make the reader read the RFID tag on the case?  These are just two examples of many details that will require consideration.  This will take time and testing to complete.

Scanners – Current scanning technology won’t read RFID tags.  It will be necessary to change the scanning technology at the front end.  For customers with scanner/scales, there may be an upgrade path to avoid entire replacments and simplify a migration from barcodes to item level RFID.  The NCR 7878OFX  has space in the unit for a future RFID reader, allowing one scanner to read both barcodes and RFID tags.  In a grocery environment, a scale will still be needed, so using the current units as much as possible would be a big cost benefit.

Security  – Using RFID tags on every item would potentially eliminate the necessity of EAS tags.  An RFID reader at the door could validate whether an item has been purchased or not when it passes through the door.  It may be more effective in avoiding false positives (false alarms) at the store exit.  There is also the very reasonable concern of privacy advocates that those with readers could read what is in your bag – or even in your house depending on the tag.  There needs to be some sort of tag destruction protocol to avoid that concern.

Like any technology, there are some great upsides, but costs and difficulties must be overcome which will drive the feasibility of this technology out for some years.  I wouldn’t expect to walk through a gate and have everything scanned automagically either – for starters, metal carts mess with the reception, tags in the middle of cases of liquid don’t always read well, and most customers won’t want us weighing every customer to see how much their produce weighs.   Given these issues and the concern around privacy (remember RFID passport hacking?) I don’t believe this will catch on any time soon.  Simplicity and cost are key and this solution isn’t there yet, but I look forward to new developments that could make it happen.

%d bloggers like this: