There is a huge amount of confusion about the various RFID standards programmes among users. Chairman of leading European Auto ID solutions enabler Scanology BV Richard Rees answers some of the most frequently asked questions


What is ISO 18000?


ISO 18000 is a family of RFID standards for use in item identification. It defines the interface between the RFID data carrier and the RFID reader or how tags and readers talk and respond to each other. There are several other ISO RFID standards used for tracking, say, animals and bank cards but only 18000 deals with the supply chain so beware.


Why is it needed?


We use standards every minute of the day. Every time someone fills up their car, they assume that the fuel at every garage will be suitable for the engine in their car.

Nobody would be too pleased if they had to hunt for the garage that did have the fuel for their car and if, when they got there, the filler nozzle didn’t fit their tank. Drivers want to be protected against putting diesel into a petrol car. Standards for the fuel, the pump and the car do this for you.

Standards may seem dull, even unnecessary, but they are what makes open system operation possible by solving the problems of interoperability and non-interference. They also encourage multiple sources of supply and innovation, encouraging the greatest functionality at the lowest cost.

Standards are particularly needed for RFID. The bar code has a pretty tight one-to-one relationship between reader and symbol. The radiating nature of RFID energy means that problems of RFID systems being able to understand each other, and not interfering with each other and/or with other users of the radio spectrum, need to be solved. Standards do this.


What is being standardised?


ISO 18000 only looks at the issues involved in how the tags and readers communicate with each other. These are high-level issues, most of which will be completely transparent to users. ISO 15961/2 were developed alongside 18000 to manage data structure issues.

Only certain frequencies are allowed to be used for RFID and the 18000 series covers all the commonly used bands.

These are:

• 18000-2 <135 KHz

• 18000-3 13.56 MHz

• 18000-4 2.45 MHz

• 18000-6 UHF 862 – 928 MHz

• 18000-7 433 MHz


Why so many frequencies? Why not just one?


It would be a whole lot simpler if there was one solution but there isn’t. The laws of physics determine that each frequency is good in some situations but not all.

Physics also affects the cost of the chip inside the tag. In general, UHF devices provide a cheap and effective solution. That’s why all manufacturers are paying so much attention to them. The interaction of UHF energy with the user environment can cause problems. Inductive systems operating at 13.56MHz and 125KHz may provide a better answer for certain applications.

Standards may seem dull, even unnecessary, but they are what makes open system operation possible by solving the problems of interoperability…”

The ISO 18000-3 mode 2 Magellan system [] is an outstanding example of fit for purpose use of inductive techniques but there are many others.


Does the use of many frequencies pose a problem?


I think not. The history of the bar code shows that application requirement and technical advance produced a range of bar code symbologies. Yet bar code readers are easily able to auto-discriminate between the symbologies to allow successful decodes.

Already, multi-standard RFID readers are available. SAMSys [] has made a strong case for this approach as a means of, not only dealing with current technology diversity, but also future proofing against [and therefore assisting] technology innovation. Initially, these readers will be multi-protocol but multi-frequency capability will materialise as user needs require.


When will the ISO standards be available?


Quite simply all the ISO 18000 parts [except part 7] are now stable documents, which means that in most cases it is possible to specify the standard to manufacturers and expect to be supplied to this standard during 2003. The final administrative approvals will be starting in mid-2003 and should be complete in early 2004.


What about other RFID standards such as the EPC Network?


The MIT Auto ID Centre program, which is developing the EPC Network, has a much wider scope than ISO 18000. It is considering all the elements needed to make up a full RFID system:

• Data content – the MIT program places a very strong emphasis on cost of the RFID chip as it has the ambition of tagging everything. It has adopted a licence plate approach – the so-called ePC – to minimise the data carried on the tag.

• Data carrier – MIT has developed some very simple RFID chips specifically designed to carry the ePC.

• Data bases – just as your supermarket checkout refers to a database held locally rather than at the head office, MIT has developed the Savant system as a means of providing local caches of information rather than having to go to a central database.

• Database search and transfer mechanisms – where to find the information associated with the licence plate data. The Object Naming Service or ONS – a variant on the DNS system which makes internet searching possible – and PML – a variant on the HTML/XML which powers internet transactions – are being developed.


Are the ISO and Auto ID Centre systems in conflict?


Politics aside, no.

• Data content – The licence plate is the same principle used by the billions of EPOS bar code transactions each day, where the EAN.UCC bar code has been used for the last 25 years to reference a supermarket database to price the product. The ePC can in fact be constructed from existing EAN.UCC data without changing functionality.

• Data carrier – In terms of function, the ISO 18000-6 data carriers do the same job as those developed by MIT. There is some debate about the difference in chip cost between an 18000-6 chip and a MIT chip but it should be remembered that chip cost is actually a small part of the total cost of a tag.

• If anybody really wants a 5-cent ePC device now, there is a proven class 0/1 data carrier operating at 428THz. Most people call it a bar code. Yes, a bar code can easily carry the ePC number that drives the MIT system!