RFID is Poised for ChangeThe prosperous RFID business is on track to grow from about $5 billion in 2007 today to over $25 billion in 2017. Without collusion, most analysts agree to figures in that region and several of them see huge volumes of extremely low cost tags forming a part of the growth - even hundreds of billions in ten years from now.
Basic rules of marketing
Firstly, selling RFID to consumer goods companies mandated by major retailers usually breaks one of the fundamental rules of marketing "Never sell to someone who does not want to buy from you". Most of the consumer goods companies in the USA see no payback from fitting the passive UHF labels mandated by retailers, indeed, they may have lost a mutual $100 million so far trying to do so, despite the RFID suppliers losing a similar sum selling tags and readers to them at a loss. The consumer goods companies are therefore quick to point out the technical problems and they use any other valid reason to delay. The contrast with the booming sectors of RFID (almost all other sectors) is stark.
Some suppliers read the market correctly and prosper
When we look at the money spent, RFID is not a UHF business, though UHF is making inroads. In standards, it is not an EPCglobal business, though EPCglobal is making inroads. RFID is a business of tagging financial, access, identification and transport cards and tickets and tagging passports, library books and other things at HF, whether we look at the expenditure on tags or on systems. It is also a booming business in tagging pets and livestock with LF tags, but LF for beer kegs, gas cylinders, roll cages, trolleys and secure access is conceding ground to HF and, to a lesser extent, UHF. There are many companies primarily in HF that have more than $100 million in profitable, growing sales including NXP, ACS, Huahong and the RFID part of Gemplus. NXP and Motorola foresaw that the Chinese market for RFID would become one of the largest - indeed, this year it is temporarily the world's largest market for RFID. These two companies, and a few others, have already established substantial sales in China.
Overselling
Although some companies that were wrong footed in prioritising UHF RFID for retailing have alleged that UHF will satisfy most RFID needs in future and HF has run out of development potential, this is far from true. To those that allege that HF is limited to one meter range and only then with a tag the size of a credit card, we must point out that DAG System and UPM Raflatac have long offered longer ranges and advances by Cambridge Resonant Technologies, in late 2007, promise a further 50% improvement. To those that say HF systems can only interrogate tens of tags at a time, we point to the global success of Magellan Technologies with Phase Jitter Modulation PJM underwritten by best-in-class-partners. Those tags can even be interrogated when they are touching.
Rapid reduction in tag cost
Then there is the issue of tag cost. Companies introducing smaller Near Field UHF tags for small items like drugs at Purdue Pharma, and for other uses, is of huge significance. By mimicking the inductive coupling of HF, Near Field UHF similarly avoids the huge problems of metal and water and even focussing of beams by curved glass that are suffered with Far Field UHF. Indeed, only one turn of antenna is needed and that makes the tag antenna cheaper than the equivalent HF item level tag. Whether the UHF chip is cheaper, the same cost or more expensive than the equivalent HF chip is obscured by the speed of EPCglobal/ ISO in preparing an equivalent EPC specification at HF and the relative pricing policies of chip makers. Neither NF UHF or HF will work beyond a few tens of centimeters but that is rarely a problem with tagging of small items.
As a result, for drug anti-counterfeiting, where the US Food and Drug Administration has gone soft on the introduction of RFID and the pressure now comes from state legislatures such as California and Florida for levels of reverse audit that are very difficult without RFID, HF and NF UHF solutions are now neck and neck. For example, GlaxoSmithKline and Pfizer have Tagsys HF RFID on their most counterfeited drugs - Trizivir and Viagra respectively.
Standards needed
Sadly there is no room for two tags on these tiny pots and blisterpacks of drugs. The pharmaceutical legislators need to agree standards for the frequency, signalling protocol and secure database and they are not doing so. They say they will let industry decide but there is no eveidence that that will happen. After all, some of the same RFID suppliers are responsible for three incompatible types of anti theft tag (like RFID but with only one bit of data) thirty years on, despite 12 billion being sold every year. Progress on standardisation is zero.
Contrast IATA agreeing one standard for baggage tags worldwide, opening up a different benefit to human safety and well being. Settling for 2D barcodes on drugs for the "mass singulation" used in anticounterfeiting "pedigree" is no answer because it is tough to automate and sensitive to misorientation and obscuration and cannot provide a good extra payback in automated logistical control. Counterfeit drugs have even been unknowingly sold through legitimate pharmacies and the problem is increasing, say the FDA, WHO and other respected observers. The equivalents to the FDA in other countries are similarly relaxed about using modern technology to tackle the problem and it may take a major epidemic of counterfeiting deaths in the West for the problem to be taken more seriously. Tragically, over 100,000 deaths yearly from counterfeit drugs in the Third World has no galvanising effect on suppliers of genuine product in the developed world.
Great importance of active RFID
15% of the expenditure on RFID, and a larger percentage of the profit, is for active RFID where 2.45GHz is increasingly the most popular frequency. For example, WhereNet, a $40 million division of Zebra Technologies, is a leader in the very hot form of active RFID called Real Time Locating Systems RTLS and Zebra has just bought Navis, a large RFID system integrator, to leverage that leadership. Heavy logistics and military applications feature large in active RFID applications, with healthcare coming up fast, partly thanks to RTLS from Cisco, Ekahau, PanGo Networks, AeroScout and others. Lockheed Martin (Savi Technology) sits astride the military active RFID business where orders of tens of millions of dollars are commonplace.
Where the action is
In short then, someone landing from outer space and wanting rapid, substantial and lucrative entry into the RFID business would first look at HF RFID to ISO 14443, which is responsible for maybe ten times the sales value of anything else. Then they would look at HF RFID to ISO 15693 with its longer range for library books, secure access and track and trace. They may also look at the boom in active RFID, mainly at 2.45GHz but with much business being done at UHF and 433MHz.
They would be amused at the frenzy of wildly different passive FF UHF tags developed for overcoming the problems of reader to reader interference, signal absorption and reflection at that frequency. With a whimsical smile they will note that the problem was actually solved another way and the best selling UHF tags employ the most primitive pattern of all because they are used on such things as A) Air baggage where Faraday cages or tunnel readers solve the problem and B) Bookshops and apparel stores where the problem is largely non-existent due to low reader density and a largely dry and non-metallic environment.
Unfinished business
A newcomer will see there is still work to be done. Airports do not want to slow down the carousels and conveyors to accommodate RFID readers that demand well spaced singulation of bags. They want to speed things up. Nonetheless air baggage tagging is a great success (headed for two billion tags a year) as is apparel tagging in Marks and Spencer in the UK (soon 350 million items yearly). NF UHF needs to be proven both practically and economically in mass markets and Metro in Germany looks like being a leader here. It has been moving more slowly and deliberately than some of its peers in other countries but this is now looking rather like the story of the hare and the tortoise.
The promised 80-99% reduction in cost of HF tags, by printing an alternative to the silicon chip, needs to be proven in mass markets. Someone from outer space would certainly be interested in the 1500 organisations developing printed electronics, including a minority even putting electronics onto paper, such as newcomer Additive Process Technologies that electroplates HF antennas onto paper using low cost materials in a reel to reel process or Hana Label in China that printed ten million HF labels with silver HF antennas last year.
However, low cost processes are useless if specifications and overheads are too expensive. Here it is interesting that there is an increasing interest in something cheaper and simpler than EPC in its currently chosen form. Some, like Marks and Spencer, opt for a stripped down specification of their own. Others are backing the Ubiquitous Product Code of Tokyo University now being trialled by governments in seven East Asian countries on the basis that you should not need to put 72,000 transistors on a gumball, a letter or even a book and the issuance of the numbers for such tagging should be very low in cost. Time will tell who is right, but IDTechEx believe the indsutry would do well to contemplate some very simple specifications - ironically those envisaged by Massachusetts Institute of Technology that started it all, and low cost administration. Otherwise its early and commendable success may be seen as but a Pyrrhic victory.
The most significant announcement?
When we look back at 2007, we may realise that the most significant announcement was that at the IDTechEx Printed Electronics USA conference in San Francisco in November
www.idtechex.com/peUSA. We refer to the presentation from Kovio, a newcomer that is printing transistors. Most of the developers of printed electronics concentrate on photovoltaics on low cost flexible substrates. This is often transparent and it will be useful for energy harvesting in the vast number of RFID sensor tags planned for Ubiquitous Sensor Networks USN backed by the Korean Government, the Holst Centre in the Netherlands, Intel, the US Military and others. Photovoltaics can be printed on top of printed batteries and the printed antenna and the printed transistors can be under them. After all, no one would put a silicon chip in a barcode or anti-theft tag and RFID intended for such huge volumes must not use a sledgehammer to crack a nut either. Another use for the new electronics would be very low cost disposable RTLS which is needed for many uses including the ultimate supply chain. Another will be low cost cards and tickets with moving colour displays.
Stay updated
To learn more attend RFID Smart Labels USA 2008 in Boston on February 20-21. See
www.IDTechEx.com/USA for details.
1.8 billion RFID tags have been sold to 2005. Key volume applications for RFID technology have been in markets such as access cards for the financial, security and safety markets, or for the automotive and passenger transport sector, with smaller markets in leisure, libraries, laundry and healthcare.
Another way of looking at the sales of RFID tags is to consider those that have a battery in them, called ‘active tags’ versus those without a battery, called ‘passive tags’. This is split as follows. Most of the active tags have a coin cell battery in them, otherwise called a button battery, and are not exactly suitable for reel-to-reel production.
Active tags 410 million (highlight car clickers)
Passive tags 1390 million (highlight cards)
The future is very different from the past
After addressing technical problems with UHF, 3.1 billion tags will be used for pallets and cases in 2006. Item level tagging (especially by pharmaceuticals) and tagging of baggage, animals, books, tickets and other non retail markets are strongly growing in value – in 2008 6.8 billion tags will be sold for such applications and 15.3 billion tags for pallets/cases, but the former tag value will be higher than that for pallets/cases.
The market for RFID interrogators will reach $1.14 billion in 2008 for EPC interrogators and $0.75 billion in the same year for other interrogators, such as Near Field Communication interrogators.
Forecasts by territorial region show that by 2010, 48% of RFID tags by numbers will be sold in East Asia, followed by 32% to North America.
Deliveries and orders in 2004 were sharply up on the year before. Even if one wrongly considers the RFID tag to be nothing more than a barcode replacement then such figures are not necessarily unrealistic, because there are somewhere between five and ten trillion barcodes printed in the world every year. However, these tags will not reach the ten trillion level before 2020 at the very earliest, where they will need to cost less than one US cent and be entirely printed, like a barcode is today.
Disruptions not trends
The curves do not extrapolate up and up. The research showed that the highest volume applications of RFID will mimic barcodes where a market for barcode labels grew then declined as barcodes were printed directly onto products and packaging. The value of that label market peaked before the annual numbers sold reached a peak. That was because of strong price erosion. IDTechEx sees the same occurring with RFID but on long timescales and with one difference. The printed radio barcodes will not use the same ink as the graphic printing in contrast to directly printed barcodes today. There will be a growing and lucrative market for electronic inks used to print RFID tags onto labels and directly onto products and packaging. IDTechEx estimates these timescales and volumes.
One cent tag controversy
IDTechEx doubt that the necessary one cent tags needed for tagging everything in the supermarket – the largest volume potential for RFID - will be profitably achieved with silicon chips within ten years if ever. It believes that giants such as IBM, Xerox, Dai Nippon Printing and Samsung that are developing “chipless” alternatives such as polymer transistor circuits and Surface Acoustic Wave SAW devices may be on a better tack for the long term.
Five cent chip tags seen as a certainty
Chip tags are certain to get down to five cents as orders approaching ten billion tags are placed. Chip tags can also address enormous secondary markets, even if Consumer Packaged Goods (potential trillions yearly), postal packages (potential 650 billion yearly) and books at manufacture (50 billion yearly) mainly take one cent and sub one cent chipless tags in due course.
Several chip manufacturers have approached IDTechEx saying that they have zero interest in producing the required sub one cent chips for five cent RFID tags but they do seek less price sensitive, more sophisticated large niche markets in RFID. The new report gives great detail on these.
Billion dollar, billion tag niches
Such opportunities run into at least tens of billions of tags yearly and disproportionately large sums on infrastructure and services. The large niches are often new RFID markets coming from nowhere, not extrapolations of past trends. They include:
- The South Korean Ubiquitous Sensor Networks USN projects backed by the South Korean government for monitoring natural disasters and for many other uses
- The tagging all patients, staff and assets in healthcare facilities worldwide for error prevention and other reasons
- Antiterrorism measures in global logistics
- Meat and livestock tagging in the face of new legislation against disease
- Tagging of high value banknotes and drugs for anticounterfeiting
Timelines are given for all of these. Many examples of legal push are examined such as the probable tagging of one billion tires yearly with embedded RFID devices that also sense pressure.
Item level coming sooner than realised
IDTechEx believes that item level tagging, particularly of drugs, will rival the output of pallet and case tagging in 2005 despite the hype about the latter programs. With the world’s leading companies such as Hewlett Packard, IBM and Samsung working on the UHF problems, they will be solved but any forecaster must reflect the fact that Gillette and others see the physics being so tough that they must redesign the geometry and materials of a significant percentage of their cases and their contents to make them “UHF” friendly while the systems are being optimised. Such nuclear options are neither cheap nor rapid.
By contrast, Pfizer, GlaxoSmithKline and other major pharmaceutical companies have decided to start tagging certain products on a permanent basis in 2005. They are not all hooked on UHF and some favour the more proven 13.56MHz waveband used with almost all item level tagging and contactless smart cards to date – about one billion items out there and working well. The range of UHF is rarely needed at item level. Indeed DHL has ordered its first million tags for postal packages and they work at 13.56 MHz.
In such a frenzy of success and failure, forecasting is a risky business. Who predicted that only modest quantities of pallets and cases would be tagged in 2004 but orders for over 150 million RFID air baggage tags would be placed? Who in the West noticed that 50 million RFID tickets were delivered against just one order in Japan? It will be interesting to see if IDTechEx has picked up once again more of what is really going on and has produced more useful forecasts in this notoriously changeable field.
![[Jerry Bruckheimer]](http://s.wsj.net/public/resources/images/HC-FO987_Bruckh_20051012184716.gif)
Click here to view the Webinar
