overview

This profile considers radio frequency identification (aka RFID) technologies, applications and issues.

It covers -

• an orientation and a discussion on this page of RFID technologies
• an exploration of business, institutional and personal applications that range from vehicle tagging and inventory management in warehouses and libraries to identification of livestock and tracking of people through subcutaneous implants
• a discussion of issues, advocacy and regulation - in particular regarding privacy.

The profile supplements the broader examination elsewhere on this site regarding privacy, the global information infrastructure, identity, consumer protection, security and e-business.

     orientation

RFID technologies have been spruiked as offering fundamental efficiencies in supply chain management, substantial benefits for agriculture and human health services, improved security and positive outcomes in applications that range from library collection management to user-pays road networks.

They have also attracted concerns about privacy and consumer protection. In some circles they have replaced mobile phones as a focus for the free floating anxieties explored in Adam Burgess' Cellular Phones, Public Fears & A Culture of Precaution (Cambridge: Cambridge Uni Press 2004). They have also been hyped as unprecedented and inherently sinister.

However many people in advanced economies are familiar with them as the basis of domestic pet identification registers, entry cards and automated road billing systems, such as the E-Tag used in some Australian tollways.

The past decade has seen significant advances in deployment of the technologies - notably integration of RFID tags with multi-user databases - and reductions in the cost of particular components. It is likely that those advances will accelerate, with a proliferation of applications, increased adoption in the public and private sectors, and debate about appropriate management or restrictions on use.

It is therefore useful to consider the technologies in their legal, commercial and cultural contexts rather than in isolation. Some concerns - and claimed benefits - are overstated. Many concerns are best addressed with reference to existing privacy principles and to application of effective protocols for the collection, handling and disposal of data by organisations and individuals.

     technologies

RFID technologies centre on

• tags (aka chips) that emit short-range radiofrequency signals
• devices (aka readers) that pick up the signal and that may be networked to databases of varying sophistication
• databases that hold customer account, stock control or other information that is used by a single entity or by multiple entities
• software that integrates information supplied by the tag to the reader with information held in one or more databases

From a transmission perspective there are essentially two types of RFID tags: active and passive.

An active tag - used in for example a vehicle traffic management system for automated payment on tollways - uses its own power source to contact the reader. That power enables transmission of a signal over a greater distance (eg up to 100 metres) than that of a passive tags. The advantage is, however, usually offset by the greater cost, size and weight of the tag.

A passive tag does not require its own power source. The tag instead derives its power from the electromagnetic field created by a RFID reader. That generates sufficient power for the tag to respond to the reader, ie to supply its information. Passive tags have a shorter range than active tags, varying from a few centimetres to a few metres and affected by barriers such as metal shipping containers or concrete walls. However, the absence of an integrated power source means that they can be smaller (one state of the art tag is as small as half a millimetre, with a 128-bit ID number), potentially more resilient and - through economies of scale - substantially cheaper.

From a content perspective we are seeing the emergence of three classes of tags: 0, 1 and 2.

Class 0 are factory programmable only. Class 1 tags can be programmed by the retailer or supplier. A proposed Class 2 will feature more memory and data.

RFID tags can be incorporated in plastic security cards, paper passports, consumer packaging or even in living creatures. Typically they are being used as unique or generic identifiers (eg for a particular card, object or class of objects). There is major commercial interest in uniform numbering schemes, similar to those used for product barcodes (and in ISBNs and other metadata), such as the emerging global Electronic Product Code (EPC) for use in supply chains. They can also be used for authentication purposes.

The tags can used in isolation - for example solely as a vehicle's registration number - or can be integrated with sensors that collect a range of data, for example temperature or salinity.

Use of radiofrequency means that they can be read in proximity to a reader and thus are not dependent on physical contact (as in, for example, a magnetic stripe card) or direct line of site (eg a laser-scanned barcode). An RFID tag can be read in the dark, in smoke, in bright sunlight or other environments where visual recognition is difficult.

Subject to interference or transmission barriers, reading can also be almost instantaneous. That speed of data acquisition is particularly attractive for supply chain applications ranging from manufacturing through to retail, with for example the diverse contents of a shipping crate or shopping basket being identified in less time (and with less labour) than is required to read the barcode on an individual item.

In discussing networking and the global information infrastructure we have noted that 'wireless' is a shorthand for applications using a range of frequencies.

Various RFID schemes thus use different frequencies that include -

125 and 134.2 kHz
433.05-434.79 MHz
1.77–2.17 MHz
2.93–3.58 MHz
7.2–10.01 MHz
13.553–13.567 MHz
918–926 MHz
2400–2450 MHz
5725–5795 MHz
5815–5875 MHz
24000–24250 MHz
5795–5815 MHz

Some of those frequencies are (or will be) shared with other devices and applications.

     what is driving development?

Development of RFID technologies and applications is being driven by four factors.

Firstly, the declining cost and increased reliability of small passive tags, heading to the point where they can be routinely incorporated in disposable consumer items rather than restricted to overall supplychain shipments (eg to identify a can of softdrink rather than a shipping pallet).

Secondly, perceptions that some RFID applications potentially have a strong business case, with appropriate ROI over a three to five year period.

Industry group EPCGlobal has suggested that US retailer Wal-Mart would save US$407 million by having suppliers attach RFID tags to all pallets, based on industry-average operating margins. Requiring an RFID tag on every item would save US$7.6 billion, primarily in labor costs associated with loading, warehousing, stock management within stores and checkout operation.

Thirdly, claims that RFIDs can leverage substantial existing investment in commercial EDI (with greater flexibility and detail than barcodes), enable dynamic pricing (similar to online airline demand-based ticket pricing) in retail environments or offer practical solutions to hitherto intractable problems such as tracking livestock from pasture to the table.

Those claims have been endorsed by major hardware, software and integration service providers - such as IBM - although much analysis is patchy.

Finally, the technology is being spruiked by particular vendors - some of which sound increasingly zany as they burn the last of their capital - who are offering solutions in search of problems.

Key challenges are -

• the cost of tags, with passive tags currently costing between US$1 and 20 cents. Large-scale commercial adoption - particular for item by item identification (rather than generically by product type) of low-value items in supply chains - is dependent on costs being reduced to a small fraction of those figures. At the moment tags typically represent upwards of 60% of implementation costs, with other spending on systems integration and on readers.
• agreement on standards for tags, devices and data (in particular numbering schemes that are sectoral rather than restricted to a particular vendor/user or that are used in most links of a supply chain)
• reluctance of potential commercial users, notably small organisations, to invest in RFID technologies and more broadly in EDI. As with uptake of barcodes, implementation is accordingly often driven by a few dominant actors in the supply chain (WalMart for example has attracted attention) or as an extension of existing EDI in particular sectors such as auto and computer component manufacturing/assembly
• optimisation of technologies (eg 'fixing' small passive tags in wet or cold environments) and deepening of knowledge about placement (eg best positioning in a crate, shipping container, vehicle, package or pet)
• addressing consumer concerns (and subsequent indecision or delays by regulatory bodies) over "the spy chip" per se or the handling of data that might be collected using RFIDs. Consumer perceptions of the balance between benefits and threats are highlighted in a 2001 Auto-ID survey (PDF).

     the media equation

Those concerns - substantive or otherwise - have been reinforced by hyperbole from advocacy groups (ie both opponents and supporters of RFIDs) and media coverage that has often been more enthusiastic than enlightening.

We have thus encountered claims that

• passive tags can "see through concrete walls"
• each US banknote features a RFID tag that readily allows the 'invisible government' (apparently a curious mix of Jesuits, Zionists and members of Skull & Bones) to track individual notes
• tagging represents the greatest revolution since the domestication of livestock, allowing seamless identification of animals "from paddock to plate"
• "RFID tags inside driver's licenses will make it easy for government agents with readers to sweep large areas and identify protestors participating in a march"
• subdermal tags will allow people to carry thief-proof credit cards under their skin
• library tags will allow protesters "to be unwittingly and remotely identified at a political rally by the book in your backpack"
• "theft will be drastically reduced because items will report when they are stolen, their smart tags also serving as a homing device toward their exact location"
• criminals will target people for kidnappings and burglaries by reading their bodies or the contents of their homes and wallets

and that

One day, not long from now, virtually any store, restaurant, or business may be able to identify you, note what clothing you're wearing--and possibly even detect how much money you have in your wallet--as you enter the establishment.

or that

Already we can imagine the likes of John Ashcroft, salivating noisily at the idea of inserting similar chips directly into the skin of every swarthy foreigner and every tofu-sucking liberal commie protester while they sleep so the government can track your movements and erase your Social Security number and stomp down your door the minute you buy a used copy of "How to Make Cool Thermonuclear Warheads in Your Bathtub." This much is a given.

The most successful April Fools Day spoof of 2004 - amusingly accepted as genuine by some pundits - was the claim that US streetpeople would be mandatorily chipped -

The U.S. Department of Health and Human Services
said Thursday that it was about to begin testing a new technology designed to help more closely monitor and assist the nation's homeless population.

Under the pilot program, which grew out of a series of policy academies held in the last two years, homeless people in participating cities will be implanted with mandatory Radio Frequency Identification (RFID) tags that social workers and police can use track their movements.

... "This is a rare opportunity to use advanced technology to meet society's dual objectives of better serving our homeless population while making our cities safer," HRSA Administrator Betty James Duke said.

The miniscule RFID tags are no larger than a matchstick and will be implanted subdermally, meaning under the skin. Data from RFID tracking stations mounted on telephone poles will be transmitted to police and social service workers, who will use custom Windows NT software to track movements of the homeless in real time.

In what has become a chronic social problem, people living in shelters and on the streets do not seek adequate medical care and frequently contribute to the rising crime rate in major cities. Supporters of subdermal RFID tracking say the technology will discourage implanted homeless men and women from committing crimes, while making it easier for government workers to provide social services such as delivering food and medicine.

Concerns have also been reinforced by inept implementation, with a furore in the UK, US and Germany for example over 'smart shelf' prototypes that covertly took photos of shoppers when a product was removed from the shelf - in the UK another covert photo was taken at the checkout for comparison, apparently to reduce shoplifting.

     futures - from barcodes to the X-net

Some predictions about the future of RFIDs seem askew.

We for example regard promo for 'intelligent' washing machines that will interrogate tags in clothing for the correct tender loving care with the same skepticism with which we treat reports on the viability of the internet fridge.

It is unlikely that EPC tags will comprehensively replace barcodes in the immediate future - even if tag costs crash - given investment in those codes by over a million manufacturers, wholesalers and retailers. Forecasts for large-scale uptake of subdermal chips also seem misplaced, with the largest market for 'human' applications in the near future likely to be collars or bracelets used by custodial and healthcare institutions.

Visionaries have pictured a world - in practice the First World, rather than parts of Africa and Asia at the far end of some digital divides - where so-called 'smart dust' provides the basis for truly ubiquitous networking. Reduction of tag costs, resolution of interference problems and major advances in data handling would permit what has been characterised as the X-net ... in which softdrink cans, woolly jumpers, mobile phones, cats, dogs, grannies, cushions, cars and potplants each have one or more tags and can be meaningfully identified through applications drawing on numerous databases.

In practice it is not enough to give each anorak, artichoke or carbon-based biped a unique number - and even a discrete internet protocol address - associated with a tag. Making sense of that identity promises to be more difficult.

    next page  (applications)