PREVIOUS	March 2004 (12.12)	NEXT

Explorations in the Globalization of Everything -- Supply Chain RF-ID Middleware, Off Shoring, Real Time Global Corporation

RF-ID and Supply Chains - An Industry Overview
We Find Three “Rivers” of the Industry Rushing Toward a Hoped for Confluence

The Nature of Organizations is Changing

Decentralized organizations, specialized and usually quite small are taking advantage of the availability of broadband, home networks, collaboration tools, wireless always connected mobility to become suppliers of goods and services within and across industries.

At the large corporate level these same technologies are being used to obtain cost cutting dividends by outsourcing everything except core business concepts. To give one example here is what Reed Cundiff Senior VP in the Yankee Group has to say about on Wal*Mart’s changing RF-ID driven direction.

The differentiator in being a company with competitive advantage is no longer one of owning the means of production. Previously when you had major manufacturing shifts of other ways of doing business what you saw was what controlled who owned the means of production and who was as a result a major market leader. Now it is more and more a case of who controls the intellectual capital that is leading to better and faster product development. There is a need to maintain a networked community of suppliers and customers in order to be able to operate in a much more fluid environment.

We are seeing RFID as the tip of the spear for Wal*Mart being able to move to a store level, consignment based, inventory model for some of their higher end product lines. This enables them to change their whole financial model. The first step in RF-ID that they are looking at is cost cutting. This will take the form of lower inventory, less handling, less shrinkage, and more efficient logistics and product flow across the supply chain.

In the future, 6 to 8 years out, it will enable them to have self-service check out and they will be able to eliminate a number of higher cost labor portions of their operations. An even bigger impact on their financial model will come from being able to enable some predefined amount of shelf space for a given supplier.

At this point Wal*Mart's whole business is dropping a big box store within a suburban location and drawing customers into the store. They will simply act as a sales channel, never owning the item. When the item is electronically swiped as part of the purchase process they will receive a distribution fee.

RF-ID and Supply Chains -
An Industry Overview
We Find Three “Rivers” of the Industry Rushing Toward a Hoped for Confluence p. 4

Wireless bar codes have been seen for several years as a way to solve supply chain and inventory problems. The have been in a process of methodical development until mid 2003 when the pinnacles two largest supply chains in the world the US Department of Defense and Wal*Mart announced that by the beginning of 2005 they would require their 100 largest suppliers to begin RFID tagging at the carton and pallet level. This appears to have set off a frenzy of jockeying among supporting actors in what had been shaping up as a more leisurely march towards new standards and networks.

The RF-ID field is chaotic and the most remarkable development in the midst of this chaos, and one that I have not seen reported in any media, is a full court press by a small group of industry insiders beginning in May 2003 and culminating in the VeriSign announcement of January 12, 2004 that it had been selected by EPC Global to run a network resolving service for RF-ID tagged goods bearing EPC Global manufacturer numbers.

VeriSign’s Public Relations staff explained to the national media that the global standards group for electronic bar coding had selected VeriSign to use ONS, a variant of the DNS registry that VeriSign runs for .com and .net, the two most important global top level Internet domains. But note that the VeriSign ONS technology was developed by the Auto-ID Center at MIT which was taken over by EPC Global in the process of becoming a new global standards body.

We have a cozy partnership with EPC Global is promoting the adoption of an EPC network that would supplant and keep in business the older network of the Uniform Code Council that has long been used to transmit and interface scanned bar code data within corporate EDI systems. The UCC and EAN (European barcodes) have formed EPC Global as a vehicle to morph a close-knit set of players and technologies into something that they hope will be perceived as an unstoppable title wave. The “solution” being offered will use old and not very secure web-based technologies to weave together what is being sold to the press as a global standards committee endorsed solution to tracking RFIDs good across global supply chains.

The components are ONS (Object Name Service). A VeriSign owned and controlled ONS registry and root that resolves queries designed to identify the manufacturer or owner of a product and enable communication between the recipient of a set of goods and the source of those goods. ONS was developed by the co-founder of the Auto – ID center Sanjay Sarma. The system will use a software tracking technology know as Savant. OAT Systems is a private commercial company formed by Prasad Putta a student of Sanjay Sarma. OAT’s purpose is to develop the commercial software that will embrace Savant, ONS and VeriSign’s global resolvers. Their hope is that their solution will become globally used for RFID tracking across company supply chains.

VeriSign explained that this was a decision of huge import putting the company in a position analogous to where its predecessor Network Solutions had been with the DNS in 1993.
My analysis of what they are doing is that they are rushing to implement a solution that is seriously flawed. For example:

1. DNS as used in web sites is an inherently insecure technology. Readers will likely remember Eugene Kashpureff’s take over of the Network Solutions Website in the summer of 1997. There is an IETF standard, DNS-sec, that is secure and unspoofable. But the standard is dead in the water and is not being used in ONS. The Product Mark-up Language to be used with ONS and Savant in a form of web services commits this leading global standards effort to the use of a public internet technology that is inherently insecure.

2. There is a second problem. EPC Global announced a likely award to VeriSign in November. Neustar was the only other serious bidder. A lengthy best-and-final negotiation ensured which would up with VeriSign as the sole awardee for the critical resolution services on January 12, 2004. It would have been technically quite possible for EPC to award a contract to both VeriSign and Neustar to run independent but parallel resolution services. On the positive side, not doing this may have simplified EPC Global’s administrative life. But on the negative side, its action ensured less competition and created s single point of failure for what could become a critical infrastructure.

3. The EPC Global architecture rests essentially on the foundation of EDI foundations of the Uniform Code Council's network for the exchange of barcode information. It is likely that the most critical determinant of success for RF-ID will be an architecture that permits continual tracking of assets in real time. What is needed is intelligence at the edges not dependent on central points of failure or control like a centralized ONS registry.

4. Finally there are serious issues of approach to authentication and authorization and the ability to deliver trust across a system that is not to be found in Verisign’s reliance of PKI. From where does the keeper of the “keys” get its authority and trust? There are points of view shared by Ed Gerck (Network Manifold Associates) and Wedge Greene on issues of security and trust that are very different from those founded on PKI.

These differing views deserve a serious scrutiny that they will not get if the VeriSign - EPC Global campaign is not subjected to very serious scrutiny.

RF-ID may be thought of as three rivers trying to get to flood stage.

The first is the Wal*Mart and DoD torrent.

The second is the UCC- Auto-ID -EPCGlobal-MIT-Verisign stampede. It is too soon to tell whether the second will become a tributary feeding throughout the supply chain into Wal*Mart and DoD . If this does not happen, then either these folk will succeed in their own vertical supply chain effort parallel to DOD and Wal*Mart or they will spin apart and fragment development.

The third 'river" is every one else. These include big companies that have not yet joined EPCGlobal. They include software providers, middleware integrators, and overall systems integrators. Accenture, IBM Global, Manhatten Associates which Bear Sterns rates as 'outperform' but whose software runs I am told only on IBM AS400 minicomputers. Expensive machines with a 20-year-old architecture. Mission Assurance, Magnugistics, EXE Technologies, SAP, TIBCO are active in RFID, while there are about ten other companies involved in supply chain and warehouse management software that are waiting on the sidelines trying to figure out where to place their bets. Similarly there are hardware vendors, chip-makers, readers and printers.
The pressure to meld the three rivers into one flood will be huge because the economic payback will be by far the greatest if a highly unified system is created. However there are many issues - security issues especially relating to ONS that must be solved before tags work throughout supply chains crossing company boundaries with ease. EPC Global is pushing "a" solution. It is by no means certain that this solution will win.

Interview with Mission Assurance pp. 10-23

In a long two part interview with Reubin Steiger and Wedge Greene, we get an overview and introduction to the events on-going within RF-ID.

[Editor’s Note: Mission Assurance uses a Service Grid approach. When we Googled for Service Grids, the best paper we found was a 2002 piece by John Hagel and John Seely Brown. See www.johnhagel.com/paper_servicegrid.pdf

Greene: You selected a very good reference. It could be paraphrasing our 1999 patent applications. Big difference is that we are not using web services and an underlying communications protocol between fixed service applications; instead, we use mobile services, floating in the grid, which are connected by arbitrary, managed protocols (web service SOAP, Secure-RMI, or light-weight protocols.)]

Hagel and Brown: “Service grids and their component utilities are entirely optional components of a distributed services architecture. They operate as optional overlays of existing networks – providers and users of web services can choose whether or not they wish to avail themselves of the existing functionality offered by service grids. In this way a distributed service architecture preserves both the simplicity and ubiquity of the underlying internet platform but offers participants additional enabling functionality if they need it. Service grids do not imply a return to the feature heavy “smart” networks of the voice telecom world. They honor the value of simple, “dumb” networks while also honoring the need of participants to access higher levels of enabling functionality through optional overlays.” p. 6

Greene: Actually, web services does support a ‘dumb network’ technology, enabling the status quo of big clients to intercommunicate via hands-distant messages. The Assurance Ecosystem is more a network resident technology built at the application layer above the network. It is ‘pervasive’ rather than ‘dumb’ or ‘smart’. This article dodges who actually maintains and owns the common utilities and how people contribute and get paid for use. These service utilities were the new version of the ‘smart network’ we were pitching Worldcom to build. Note also they advocate switching burden from edge to these utilities.

[Snip]

Steiger: We are right now at an inflection point in the technology awareness curve with RFID. You have a situation where the development of standards within the global infrastructure of electronic product codes is still ongoing. But at the same time you have two of the biggest organizations on the planet Wal*Mart and DoD moving full speed ahead to adopt. The fallout of this will be that other large players will find themselves dragged along.

[Snip]

Steiger: Let me try to draw a lot of this together. Where we are then is that we have a real incentive and a movement that has a lot of momentum behind it. The goal is to apply wireless bar codes to physical inventory so that you can do things that save you money. Immediate cost savings will simply be labor savings from eliminating hand scanning. There will also be a reduction in rates of theft or more generally “shrinkage”, though this has not been quantified yet. Companies will also be able to hold less inventory because they’ll be more efficient and able to route things more intelligently and maintain physical and financial ownership of a smaller proportion of their total global goods base. The company that does this can expect to gain a benefit proportional to its size. To sum up: Early ROI for deployment of RFID mostly comes from labor saving in the elimination of the hand scanning of barcodes and its replacement by automatic scanning of RFID tags.
One question that arises in this context is why don’t existing data interchange methods evolve smoothly into this new “paradigm?” The answer to this we believe is an issue of complexity and scale. You are talking about the need for systems to deal with probably two orders of magnitude more events than the corporate supply chain systems now in use are able to handle.

[snip]

Perishable Product History Supply Chain Scenario

Greene: OK. In any supply chain there is manufacturing information that is important to be tracked and incidental to the actual tagging. The supply chain is going to involve the product itself that is going to be manufactured and altered and shipped. As components it has assembly, packing and unpacking, transportation and finally receipt and shelving at a retailer. In some circumstances it can also have returns that can go backwards within the supply chain from the retailer back to the manufacturer. There are two things that we are going to consider within this. One is an RF-ID tag that I want you to think of as a repository of location information. When tag is read, you get association of the item with the location at which it was read and the time of the reading.
An RF-ID tag is essentially a kind of location service. In addition to this it is important to understand that a fully integrated and supported supply chain will also have sensors. Sometimes these sensors are independent of tags and sometimes they are actually attached to expensive tags. Sensors monitor characteristics of the environment. For perishable goods the ones we are going to concentrate on are temperature and time.

Let’s start with a perishable supply chain. In this case a food animal like a lamb or cow. They will record the farm from which is comes. They will record the kind of chemicals being used in the farm environment. They will record the injections. If it is an animal like a cow, it will likely be RFID tagged from birth itself. People will use that ID to record what happens to that particular animal in terms of shots, of what it ate and how it got to a feed lot, what it was feed when there. This will be information that was gathered in the manufacturing part of that supply chain.

You will then get to the point where they ship the animal to a slaughter-house and at that point, if not before, they will write an RFID agent with the aid of our system. The agent will be encoded with the tag ID. It will have been started early within the lifespan of this animal. The tag then will have recorded in it all the data on where the animal was raised what it was fed, what kind of shots it got. When they ship the animal to the slaughterhouse, through our secure network the “sidecar” agent will be passed to the slaughterhouse which very likely will receive the agent before it receives the cow, because when the itinerary of where the tagged animal is being shipped, the electronic transfer of the information happens very rapidly – much faster than the physical transfer of the animal itself. There will be an Ellipsis System RF-ID agent waiting there for the tag to come in and be read. When the it comes in and the tag is read, our system associates the agent back with that tag and then it will begin to trigger policy.

Off Shoring, pp. 24 - 28

COOK Report: Where did these processes get started and how did they evolve?

Vashsitha; Initially it started with Ireland. But it has gone way beyond Ireland, which suffers from a labor imbalance. They cannot produce as many graduates as are required to keep this market growing. In the meantime other countries have learned from Ireland and have larger labor force and an infrastructure that can sustain growth. When you look at Bangalore and Hyderabad in India; Shanghai, Beijing, Shenzhen in China and Manila and Cebu in the Philippines, you realize that very similar factors are at work. Now some countries are ahead of the others in certain areas. But the factors are generally the same, education systems, government regulations, infrastructure, focus on process and quality, the presence of multinationals, venture capital - the confluence of all these processes is what's creating the offshoring market.

At the same time when you look at the buyer markets that are the UK and the US, what is determining that the time is right for these processes to take off? The fact is that people are ready and that the technology is there to drive what you call the real time corporation and enable the encapsulation and transport of business processes. Also driving this is the need for American companies to be fast to market and in the past two years to be much more cost effective.

When you look at the kind of work that is going off shore you will see that leading the way are the commoditized labor processes. Low end simplistic and repetitive. Let's look at the jobs that are migrating today realizing as we do that what will be leaving 3 years from now and five years from now will be very different on top of what is happening now.

In looking at the exportability of work, I would break it up into three segments. For things like business processes, first would be things like data entry, the next would be processing and the most complex would be analysis. I can talk about applications in very similar ways. If you look at applications on the data entry side is technology that has been there for a long time and for which you don't need refined skills sets to process. For example html is old technology. In medical records the data entry is the low end side whereas the development of medical records technology is still on the cutting edge. If you look at market adoption, you will see that electronic medical records have not yet been widely adopted.

Contents

Supply Chain RF-ID Middleware, Offshoring, Real Time Global Corporation -- Explorations in the Globalization of Everything p. 1

RF-ID and Supply Chains -
An Industry Overview
We Find Three “Rivers” of the Industry Rushing Toward a Hoped for Confluence p. 4

Enterprises Investing at the Edges
Trend Necessitated by Increasing Importance of non
Vertically Oriented Supply Chains p. 7

RF-ID Middleware
An Introduction to the architectural concepts, software supply Chain Virtualization and Organizational and infrastructural players p. 10

Offshoring
An Entire Industry Grows Up Designed to Facilitate Export of Jobs -- Telecom Liberalization in India China and Elsewhere Creates Infrastructure that Enables Foreigners to Become US Telecommuters p. 24

Coming in our April Issue a major interview with Art Kleiner on Core Group Theory, one with Keith Dierkx and one with Brough Turner on Supply Chain issues and a private mail list discussion on real time global topics.