Using RFID to merge information with physical items in order to enable new business processes

RFID (Radio Frequency IDentification) tags present a new way of being able to track and identify physical items.  While RFID has been in use since World War II, it’s application in retail and supply chain management has been slow in implementation.  There are many types of RFID technologies, but they all follow the same concept.  Some sort of electronic tag is affixed or embedded in a physical item and this tag can be remotely sensed to gather its information.  Most of the tags we see today are passive, in that they do not themselves broadcast their information, but react to a reader machine that interrogates them wirelessly. 
Many people have suggested that the next major phase of the Internet will be in the integration of digital information and physical things.  Chris Anderson describes a “new industrial revolution” as being what happens when the Internet does for physical goods what it did for non-physical goods, like music.  Kevin Kelly describes the new economy in the next 5,000 days of the Internet as being tightly integrated with the physical world.  Ray Kurzweil’s “Singularity” takes this concept to its furthest extreme by proposing that physical things, information, and humans will eventually merge into a single entity.
RFID is one of many technologies that can be used to merge the two worlds of physical networks and information networks.  An RFID tag on a physical object can be that sliver of information in the physical world of which Kevin Kelly speaks.  This merger of information and physical objects could radically change the way we go about sorting and cataloging items in shelves.  Today most stores follow a classification system similar to what libraries follow.  Items are classified as a certain type and then typically organized on shelves with similar classifications around them.  For example, you’ll find cheddar cheese where all the cheeses are grouped.  One fundamental limitation of this type of organizing is that an object can only be in one classification group or another.  In a library a book is either classified as law or humor because the book can’t physically reside on both shelves at the same time.  So, if you find a book on funny laws it’s difficult to classify and place.  Do you put it in law or humor? 
The Internet gets around this problem by not placing things in exclusive categories.  Information ‘objects’, like websites, are described with meta information such as tags.  With these searchable descriptions a book can be both humor and law (and many other things).  If we apply this concept to the storage and retrieval of physical objects, then the organization of the objects would no longer depend on their physical location.  How would we do that?  First, let’s take a look at how one company already uses RFID.
One retail company that has heavily invested in RFID technologies is American Apparel, who’s RFID based process for handling inventory has four basic steps.   First, the product is delivered from truck to a receiving station where it is scanned and placed in a stockroom.  Next, product is sent to fill station to be scanned before it moves to the retail floor.  The third step is a validation point where the product scanned to validate that the employee has all the correct items before they are placed on sales floor.  The last step is at the point of sale when the product is scanned just before it leaves the store.  The scanning of the RFID tags occurs at immobile stations or with mobile handheld scanners.  This process seems to have gained their company much efficiency, yet it seems even this implementation lacks the imagination to utilize the technology for its enabling potentials. 
When you look at this process you can see a few things.  First, it looks a lot like they are taking an old inventory process and just making it a lot faster with a new technology.  This is great, however RFID can enable new and more effective processes, as I’ll explain soon.  Second, a lot of the extra steps in their processes have to do with a deficiency in their implementation of RFID.  By using hand scanners and only a few stationary scanners the store can know what it has at all times but it still has an issue with knowing where it’s goods are.  This is a subtle, but important point.  American Apparel’s processes work around this lack of location awareness by implementing extra steps. 
Accurate information on where inventory is at any given moment could be obtained by integrating the RFID readers into the shelves and storage units.  The shelves could then wirelessly report back to the inventory system which products they have.  When a product is moved from one shelf to another, that location change would be picked up immediately. 
If a store can know at all times what it has and where it has it, then you only need two steps in your inventory process.  One step for taking products out of the truck and putting it in the store (it doesn’t matter where, because it’ll be accounted for and located by the RFID readers integrated into the shelving) and another step at the point of sale to let the system know that the item has been sold and not lost or stolen. 
By knowing what product you have and where it is, you can separate the tasks of inventory management and sales.  Separating these two tasks gives the sales staff the freedom to rearrange the store displays in any way they like and at any time.  It allows them the freedom to make the most effective arrangements of product for the current sales and marketing conditions without concern that their actions will cost them hours of inventory labor at the end of the day.  With this new process, one could even make a reasonable argument that store stock rooms may not even be necessary anymore.  After all, wouldn’t this space be better served as retail space for displaying the goods to customers? 
By merging information networks with physical goods we can enable entirely new processes for businesses.  It would be a shame to see American businesses use a technology like RFID to simply make existing processes run faster.