Can read/write UHF transponders work reliably?

The true test for a new technology is will the device work reliably in a real world situation. Even the current read only UHF technology that is being tested in supermarkets at present, is finding a problem with its system reliability, with reading errors causing problems in the automated system. These errors are typically caused by transponders not receiving enough power to operate when passing past a reader, caused many times due to the transponder being screened by other goods on the pallet, or being incorrectly attached to an item in such a manner that the item sucks up the available RF energy.

To store information on a passive tag, certain memory structures are needed which can be programmed and which will retain their data even when the energy is removed. This is achieved by storing charge in the semiconductor layers in the chip where the leakage is very low. To achieve this storage, a higher operating voltage and energising power is needed, compared to those needed for just reading or operating the device. For example an EEprom can be built using 2.7 volt architecture, while a read only device can be made at voltages as low as 1.1 volts. In addition the power needed by the transponder is higher for the writing process, compared to the minimal power needed for the reading cycle. There are pumped storage devices that can increase the voltage available for programming, but they require higher powers for operating and starting. The transponder also cannot have any significant energy storage on the chip to cater for these higher energy requirements due to the goal of what is achievable in producing capacitance while wanting a small chip area.

The reader radiates power from its transmit antenna. This energy is spread over the surface area of a sphere whose radius is the time of travel times the speed of light. This results in the energy density available to the transponder decreasing as the inverse square of the distance, - that is everytime the operating range doubles the energy available at the transponder is quartered. This property means that transponders operating at lower powers will have greater operating range than those needing higher powers. This further means that due to a UHF transponder needing higher power for the storing of energy than that needed for reading, the zone in which it can update information (write) is going to be smaller than that at which it can be read (possibly as little as only 30% of the read zone). Further as it is not possible to determine in which zone the transponder lies due to the high speed of light and the small operating bandwidths, the reader is not able to determine accurately if the information it is wanting to update will be successfully implimented.

In our newsletter we discussed the issue of reflection of energy off interfaces between objects of different dielectric constant. The reflected energy from these interfaces interacts also with the energising field from the reader, causing RF "holes" even in the read/write zone, meaning that even in the read/write zone there will be areas where updating information on transponders will be unreliable.

The test for the success of this technology is going to be real world testing. This can only be achieved when large numbers of the transponders are in use and when system reliability can be accurately measured.

Possibly the next EPC phase would be to go to a protocol which impliments a number system based on the current barcode systems so that barcoding and RFID devices can be used together allowing the smaller retailers to also benefit from the new wave of automation!

(N.B. It would also be a good time to switch to Tag-Talks-First type protocols so as to minimise the RF interference levels and make the new systems compatible with the limited RF spectrum available in most countries!)