PDF File: 

ASC6_wp_gibbs.pdf

Intel Corporation

Modernizing supply chain solutions through the convergence of computing and communications technologies.

The buzz around radio frequency identification (RFID) as the new hot technology is almost deafening. It’s not uncommon for the computing and information technology community to hype new technology. It happens with regularity, but it’s rare when multiple industry sector leaders and government policy makers join in the fray. And that is clearly the case with RFID, which is emerging as a core subject in industry conferences from retail and consumer packaged goods manufacturers, to health care, manufacturing, pharmaceuticals, energy, and more recently, the Department of Defense. It’s hard to find an industry sector where RFID isn’t one of the top issues being discussed.

RFID is the application of a miniature radio transmitter with logic and memory. Imagine a bar code tag with brains and the ability to send signals as well as read them. The concept of integrating radio frequency transmission on a chip has been around since the early days of the semiconductor industry, so why is it hot now? A number of factors are conspiring to drive the excitement, including the cost of the chips themselves, the emergence of ubiquitous wired and wireless networks, and new software technologies coming to market that allow distributed systems to work together. The current capabilities are being complemented by some exciting research and development just over the horizon with sensor area networks, high bandwidth wireless communication standards that work over very long range, autonomic computing architectures, and agent-based software systems.

The cost and capability of RFID chips is the most pronounced of the recent events. The price of an RFID chip that can contain 128 billion bits of information and send a signal roughly 3 million miles today, crossed the threshold recently such that it is now reasonable to put the tags on palettes and containers and apply readers in standard distribution and shipping environments. The dynamics of the semiconductor industry suggest that it won’t be long until any manufactured item can be tagged and the transmission distance will be close to 30 meters.

The Law

Dr. Gordon Moore, the previous chairman emeritus and cofounder of the Intel Corporation, is famous for Moore’s law, which states that the number of transistors that can be manufactured in a semiconductor die doubles every 18 months. The law is based on empirical observations he made in the late 1960s and it’s truly amazing that it has held true for the past 40 years. Current best estimates are that it will continue for another decade. The staying power in my mind has been as much a law of economics as one of technology. The semiconductor industry invests $10 billion each year in R&D, and this investment leads to the innovation spiral. Over the past few years, Intel has been investing $4 billion to $6 billion per year on R&D, and based on this investment has been able to move the bulk of its product manufacturing to 90 nanometer (nm) process, and is heading toward 65 nm in the near future.

Moore’s law is at the core of the new RFID phenomenon, where two things are guaranteed: The RFID chips will get smaller and cheaper; and as the technology gets more reliable, standards will emerge and proliferate in ways that could only have been imagined 30 or so years ago when early RFID pioneers started to experiment with the technology. Dr. Moore’s observation of a generation ago makes it a certainty that it’s only a matter of time until the full promise of the technology is realized.

While the technology itself is exciting, the economic impact of how it can be used is what’s driving the entire buzz. Dr. Moore made another lesser-known observation, which was that the technology used by the majority of businesses as a recession is entered will be different from the technology used as the economy recovers. The observation is interesting, and it was largely based on the economic history of the past 30 years and was focused on the information technology industry. However, a cursory look at economic history since the beginning of the industrial revolution indicates he might have been onto something very powerful.

One can lay out the milestones of technical innovation and map them against economic cycles from the advent of the industrial revolution and see a fairly obvious sequence: textiles, railroads, telegraph/ telephones, automobiles, computers, and it appears that nanotechnology is emerging as what comes next. A common example that is often pointed out at this level is the link between the ramping of the railroad to the explosive market growth between 1840-1860. Within each of the macro technology trends there are sub-trends. Here often-cited examples are drawn with the automobile industry where mass manufacturing led the boom following World War I and mass customization led the boom following World War II.

It is the subtrends within the computer industry that Dr. Moore was referring to, and one can see obvious mappings from the mainframe, to the minicomputer, to the PC and the Internet. As the world wide economy digs its way out of the downcycle of 2000-2003, wireless high bandwidth communication networks converging with distributed computing systems appear to be the innovative macro technology trend that is emerging. As part of the overall wireless computing and communications trend, RFID and the related technology of sensor area networks that serve as an interesting bridge between the computer and nanotechnology, promise tremendous potential for multiple industries to modernize business process and get one generation ahead.

Why All the Hype?

Applications are well under way in a number of industries. Avis, the global rental car company, uses RFID in fleet service management. In this application they equip each of the rental cars with RFID sensors that are linked into a service and maintenance database. Following each rental, the RFID tag is read automatically and based on mileage and time of service, the system automatically updates the records and sends a message to the service personnel to perform the appropriate maintenance.

Early implementations such as the Avis example indicate that the technology is ready today and can be used to gain a business advantage, but the hype is driven by the potential of RFID to change the game for logistics and supply chain management. Leaders in multiple industries in the public and private sector expect the potential savings in operational costs and improved customer service to be in the billions.

For example, consumer packaged goods leaders estimate that out-ofstock items cost them more than $1 billion annually. This is the benefit applied to only one business function. If one goes beyond this and looks at potential improvements in inventory turns that can be gained through improved forecasting made possible by greater knowledge of where goods are in the distribution chain and what is happening at the point of sale, then the gains are even greater.

If you take a deeper look at the timeline mentioned earlier, one of the motivations driving these historical technical innovations was the desire to manufacture goods faster, more cheaply, at higher quality, and then move them as fast as possible to the ultimate consumer. The excitement with RFID is the ability to apply autonomic (no human in the loop) tracking of goods from manufacturing through sales and in some cases postsales, and apply new business processes to take advantage of the improved information accuracy.

The previous technologies all had one glaring limitation that RFID is intended to solve: accurate knowledge of the goods’ location, particularly as they near the point of manufacture/sale. The knowledge of where the goods are in the supply chain from manufacturer to distributor vary, and some have used bar codes and some fairly laborintensive business processes to develop reasonable accuracy here. RFID approaches offer the potential to save significant labor cost during the middle stages of the supply chain.

The critical information gap occurs on either end, between the times when the goods leave the distributor’s truck and when they are actually built or sold. This is often referred to, tongue in cheek, as the last 100 meters. This is where the hype is focused.

Simply having the information is important; how it is shared and with whom will dictate the ultimate economic benefit. The business problem surrounding this issue is so pervasive that nearly all M.B.A. programs consider it in one way or another. A popular approach is MIT’s Beer Game, where students take on the role of a small convenience retailer, a beer distributor, and the brewer. In the Beer Game, the players are all given the kind of incomplete information that is typical in a real supply chain. Human nature then takes over to highlight what problems can occur, where small variations in actual demand – the amount of beer sold or not sold – leads to huge swings in actions taken by the distributor and even larger swings for the brewer.

The swing in over/undersupply is commonly known as the bullwhip effect, and the cost in increased inventory, labor to manage it, and potential lost sales due to underforecast is what drives the billion dollar potential. The potential for RFID, along with the right IT infrastructure and business process to solve this problem, is what is driving all the hype.

Intel and RFID

Currently, Intel has no plans to build RFID chips. Beyond that we are active at multiple levels, from building blocks for readers to Itanium-processor-based servers, to solution services practices which focus on adaptive architectures that help enable inevitable changes in business process. We also have multiple R&D efforts in sensory networks that include work with robotics, motes, and smart dust.

Our primary focus is to stimulate new usage and value for core IT infrastructure building blocks from the appropriate adoption of RFID, including servers, communications networks, desktops, mobile laptops, and handheld platforms. We also want to make sure that our long-range road maps comprehend the new usage as it materializes so we can help ensure the right product segmentation and features.

We have a number of external labs under way in RFID and related sensor networks, motes, and autonomic computing. Intel IT has already implemented wireless networks in the fabs to help automate workflows, primarily by taking paper out of the process and speeding up communication to key decision makers. We are exploring RFID and sensor networks to further improve workflows by improving the management of work in progress and to monitor atmospheric conditions in the fabrication facilities.

The solutions market development group has been actively investing in prototyping, early proof of concept, and standards definition. Key among these is our investment in the Auto-ID center at MIT.

What You Should Do

RFID is an exciting component of a larger industry trend toward the convergence of computing and communications and pervasive, or autonomic computing. Within this macro trend wireless communication is emerging as a critical communication model that is required to derive the ultimate value since it untethers the users and the systems, and RFID is seen as the component that will close the gap of the last few meters.

Within this trend are usage models that will improve the efficiency of current business process that are generally viewed as productivity enhancement. RFID, and the wireless computing/communication trend it is a part of, also offer the potential to gain new levels of value by introducing processes that don’t require humans in the loop at all.

While some of this may seem futuristic, and some of the visionary ideas are certainly beyond a five-year horizon, early pilots indicate that the technology is robust enough to be deployed today, from traffic control, to portions of the CPG/retail supply chain, to health care. Perhaps more interesting than the fact that all this stuff works, is the fact that it delivers significant business value.

Any business with a supply chain should take the following steps:

• Get smart on the technology. It’s not rocket science, although it’s close, and it’s important to understand what it is and isn’t all about;
• Benchmark your current supply chain and understand if/where having better information about inventory levels, workflow, and forecasting could drive business benefit;
• Determine if there is a process that would benefit from using RFID, do a rough ROI analysis, and if the indicators look positive, then kick off a pilot; and
• Don’t wait, and don’t limit your thinking to mandates or simple compliance.

Conclusion

There are gains to be made today, so don’t sit on the sidelines. You should heed history that shows the average timeframe for an industrial innovation’s maturity rate as more than 20 years. You should expect that RFID won’t happen overnight. It will take time for the tag prices to come down, the infrastructure to use the information to be implemented, and the appropriate business processes applied. In the end the market will grow, and technologies like RFID will be one of the reasons why.