RFID Reader Deployment FAQs Volume 1

June 11, 2011

In this first post of frequently asked questions regarding deployment of passive UHF RFID in challenging scenarios, I will answer two questions:
Q1. What are the current countermeasures to combat reader to reader and reader to tags interferences for UHF RFID?

Q2: What would be your advice for best practices while installing UHF RFID system in the environment full of metal objects?

A1. As many of us experienced in actual deployments of RFID systems are aware, there is a difference between how one reader performs in an environment with all variables being controlled (movement of forklifts, location of tagged items, etc.) and the “real world environment” where none of these are controlled.

For UHF readers which are EPC GEN 2 compliant, there is a Dense Mode of operation which should allow for numerous readers in an environment. However, what has had to happen in the past is that the installer or integrator has to configure for worst case and give up a lot of potential performance. In a reader, configuring for worst case means selecting one of the over 256 available configurations (reader modulation, pulse width, tag to reader link frequency, Pulse Interval Encoding ratio, etc.) or, more typically, just going with the reader’s default (static) settings which are, again, usually configured for worst case.

The flip side is to optimize configurations for highest throughput based on the given conditions (usually controlled) in a pilot test (usually a single or a few readers) which works great, amazing throughput/tag read rate – the installer collects their check and moves on. Now, the environment changes – metallic shelves are installed or moved, forklifts are moving in and out of read zone, overhead lights are turned off and on throughout the day, more readers are installed…you get the picture.

Enter the Impinj Speedway Revolution reader with the Autoset feature – the ability to monitor the environment as well as connected antennas and tag reads to optimize configuration of all these variables in real time.
You can see an example here:

Q2: What would be your advice for best practices while installing UHF RFID system in the environment full of metal objects and very high humidity?

A2: I shiver at UHF RFID installs which are full of metal objects because it leads to a highly changing RF environment (those RF signals bouncing everywhere!); sometimes it works in the favor of reading desired tags and sometimes not. What usually happens is that an installer configures for worst case reading at a portal or chokepoint which usually means higher power and sensitivity which leads to stray reads due to reflective nature of metal in the environment.

The answer here is a combination of the AutoSet feature so that the reader can adjust reader and tag communication settings according to the environment but also zone control – defining a read zone and eliminating strays. UHF RFID has gotten really good at reading tags, often 99.x % performance, what happening now is reading too many tags! That is, reading tags outside the desired read zone.

Of course there are the usual ‘knobs’ on the reader to adjust (transmit power and receive sensitivity) but the often-overlooked additional factor is reader antennas with narrow beamwidth to create constrained read zones.
If only a limited range is required (i.e. < 50cm), you can use UHF near field tags and reader antennas like the Impinj Brickyard or Mini-Guardrail:

For longer range, like in a dock door application, a typical circular-polarized patch antenna has a 6dB beamwidth of 90 degrees – this means that is a tag is located at a 45 degree angle on either side of it, it will be read at around half the maximum distance.
For example, if you have a highly sensitive tag (say, one using Impinj Monza 4 IC) and are using maximum allowed transmit power with highest reader sensitivity setting on the Impinj Speedway Revolution reader, and you find that the maximum read range directly in front of the antenna is 16 meters, then when tags are off at a 45 degree angle, the read range will be 8 meters, not a very constrained read zone!

So, you might want to consider something like the Impinj xPortal which uses Dual Linear Phased Array (DLPA) technology and has a horizontal read zone of only 60 degrees (3dB).
If you require an even tighter beamwidth antenna, there is the MT-263020 from MTI Wireless which has only 30 degrees beamwidth.

Lastly is the option to use some shielding material such as RF absorbing foam or cloaks and adjusting antenna angles – the problems with these two approaches to constraining a read zone are that, especially in an industrial or warehouse environment, the shielding may get damaged or destroyed and antennas get knocked or bumped out of alignment (or the retaining clamp simply comes loose); so, as much as possible, use hardware and configurations that are not too dependant, if at all, on these.

As always, any feedback from readers is encouraged.

Determine RFID Tag Direction and Velocity using Phase Data

June 3, 2011

It is not a new concept to use the reported phase from a UHF GEN 2 RFID tag to calculate direction of travel as well as velocity. Phase data can also be used to filter out stray tag reads, assuming the stray reads are static/non-moving tags while desired tags are in motion (i.e. through a portal), using an approach known as TD-PDOA (Time Domain Phase Difference of Arrival).

Here are a couple of resources to help you with implementing this technique. The first is a white paper which gives an overview and some theory on this approach and why it is the best versus some other options:
Phase Based Spatial Identification of UHF RFID Tags

Next is an application note from Impinj which gives some practical specifics on using the Speedway Revolution reader to get the phase data from the tag and how to use it to calculate velocity (and thus direction of travel):
Use of Phase Data with Impinj Speedway Revolution

The latest version of firmware for the Speedway Revolution reader from Impinj will also allow you to read out doppler frequency directly. However, you actually need to slow down the read rate such that using that particular approach to measure tagged items moving at a walking or even forklift pace is not very practical.

Remember that phase resolution and time window are inversely related; you need a very long time window (packet length) to get down to the few Hertz of doppler resolution required at UHF frequencies and typical velocities. So, for most applications, utilizing phase data is the best approach.