Although it may seem like there is a lot leading up to the actual on-site deployment of an RFID network, the up-front work can really pay off when you are on-site. My father used to say, "Measure twice and cut once." This is from a guy who had a circular saw that for at least 15 years never made it out of the box. Nonetheless, his advice has stuck with me.
The work getting the site analysis, creating a POAM, and putting communications standards in place will all meld together to make sure that deployment goes quickly and without surprises. The following sections will give you the step by step process to install the RFID network.
The site analysis and design document are your blueprints to start building your RFID network. Ideally, you should use a tool such as Microsoft Visio to put together the design of each portal location so that you and anyone else on the installation team knows what the final installation will look like.
Take a look at Figure 7.3 for an average dock door layout.
Figure 7.3: Visio diagram of an average dock door
It is always a good idea to get on-site a day or two before the actual equipment arrives. Bring your blueprints or your deployment software and make sure all the portal locations that you laid out in the site survey have everything they need for proper installation of the portals. The big things to look for are as follows:
Proper electric circuits dedicated to the RFID reader
Adequate grounding for electric outlets
Local area network (LAN) connection punch-down
Live LAN connection or WiFi connection
Clear, nonobstructed installation space
A ground loop is caused by an unexpected current that flows in a conductor and connects two points that are technically at the same potential, but are actually at different potentials. Because the potentials are different, this causes one of the points to be out of balance and usually results in some circuitry being overloaded and burnt out.
Refer back to Figure 7.3 and you'll notice that there is a clear representation for the electrical and LAN connections. After the site survey, you should have a good idea of where each dedicated 110 V circuit for the RFID reader needs to be to make the deployment successful. If you can standardize the location and create a very clear layout, similar to Figure 7.3, you will save a lot of time, money, and headaches.
The diagram of the portal layout should be distributed to the site electrician who can then know exactly where to put the electrical or LAN connections. The first thing you do when you get on-site is to make sure that these are in the proper location and are functioning properly. Testing power is easy: plug something in and turn it on, or use a circuit tester. The LAN requires a little help from a network analyzer or special LAN connection testing device. Either way, make sure you test connectivity as one of the first things when you get on-site.
Different objects can possess a property known as electric charge, often manifested as electrostatic discharge (ESD). What happens at the time of discharge is that an electrical field exerts a force on charged objects, accelerating them in the direction of the force. This happens, for example, when you drag your feet across the carpet in the dry winter months and then touch your cubicle mate's earlobe with your fingertip. This force has the same direction as the electrical field vector (pointing your finger), and its magnitude is given by the size of the charge multiplied by the magnitude of the electrical field (how much you scuff your feet). Touching the unsuspecting earlobe results in discharge. You need to make sure that you don't discharge any current on a tag or reader, rather on a grounding device or ground strip.
Most people deploying RFID racks never give thought to where they are being installed. The same can be said for the folks designing the racks that hold readers and antennas. Two critical things to keep in mind are the weather just outside the dock door and the infrastructure supporting the facility.
The weather pertains directly to wind or rain impacting the reader or antennas or lightning affecting the facility. A properly designed and installed RFID rack should have a flange of heavy gauge steel welded to the bottom of the rack, which can be bolted into a concrete floor. This should be immediately bolted to the floor in its proper location.
The next order of business is to cover all the fragile contents of the RFID rack such as reader, antenna, PLC, light indicator boxes, and wiring. This is best accomplished with material such as Lexan, which does not affect the RF radiation pattern emanating from an antenna. Although the rack should also be vented to allow for escape of the limited amount of heat created by the reader, more important is that the entire enclosure be covered so that dust from the warehouse environment does not enter the fragile inside of an RFID reader. Figure 7.4 shows the setup of ODIN's hardware-neutral RFID rack with one of the Lexan covers off the side.
Figure 7.4: RFID rack with side cover removed
The other likely problem to be caused by nature is a power outage. If you look at the rack in Figure 7.4, which is ready to be shipped to the client's premises, you'll notice a cardboard box in the bottom. Inside there is an uninterruptible power supply (UPS), which provides protection from power spikes, ground loops, and brownouts. A simple UPS that can keep the reader running for 20–30 minutes will ensure that configuration files are not lost, and data is not missed in the event of a power hit. The other component of this system is making sure the portal is properly grounded in case there is a lightning strike that is looking for a place to ground out.
System installation accuracy is critical to meeting client objectives. You can ensure that the RFID reader configurations are accurate for all instances by measuring the RF performance physics, in other words how do the various radio frequency waves behave as they propagate through out a facility.
There are a several tests that help you optimize an RFID portal, namely critical boundary, RF radiation, power levels, and other settings crucial to achieving 100 percent read rate. In addition to achieving 100% read rate, it is important to remember you are trying to do just the opposite of setting up a WiFi point (which tries to flood as much area with a signal) you are trying to limit the broadcast area so you don't create any ghost reads, which are reads that aren't actually taking place at the portal you are testing. This gets down to limiting interference in the multi-reader environment. ODIN technologies offers an automated services delivery software called EasyReader which analysis and final configuration specification is automatically recorded and can then generate "as-deployed" configuration reports. If you don't have a access to EasyReader what you want to maker sure you do is follow a process and document all the steps in that process.
The process should be:
Install and connect to the reader.
Set the power level for each antenna for 100% read success of your product.
Layout a boundary where you do not want reads to take place, the critical boundary.
Reset the power down to a level that does not give reads outside the boundary.
Using a spectrum analyzer measure the level of power at various points inside that interrogation zone.
Record the settings and levels for the configuration database (Microsoft Excel works well for this).
Many companies new to RFID will configure using a trial-and-error approach. This is a recipe for disaster and lengthy deployments. Using a set process or an automated services delivery tool is the right way to ensure repeatability and accuracy.
The site assessment will allow you to determine the maximum number of configurations. So, for instance, you should be able to come up with two or three baseline configurations such as a standard dock door or a personnel door. You can preconfigure your readers with that setup based on where they belong. This is also why having that IP address for each reader and each location is so critical, so you can preload the configuration.
Here are the typical steps for configuring a reader:
Step 1: Connect antennas to the reader. If it is a bi-static antenna, make sure that you connect the transmit (Tx) on the reader to the transmit on the antenna, and the receive port on the reader (Rx) to the receive port on the antenna. (The ports on the antenna are not usually dedicated to transmitting and receiving; this is determined by the connection to the reader ports.)
Do not power on the reader without a proper antenna connection or other form of protection on the antenna ports (terminators). Doing so could overload the reader and fry the unit.
Step 2: Power up the reader. Most readers have a green power light indicating that the reader is turned on. Usually during the bootup cycle, the green light will flash. Before you proceed, make sure the green light is steady. If it continues to flash after power up, there is likely a firmware problem and the unit will need to be checked for its firmware version or replaced.
Step 3: Set up the direct network connection. This is usually referred to as a crossover. The crossover allows you to communicate to the reader directly from a laptop.
Some PCs or laptops perform the conversion by themselves so that you will not even need the crossover cable, but a regular Cat 5 or 6 will do.
Make sure you keep in mind that a regular Cat 5 or 6 cable (the type of cable that is used to communicate over a local area network) is different from a crossover cable.
To set up the network connection, follow these steps:
Connect the Cat 5 crossover cable to the reader and to the host PC or laptop
Configure the network interface on the host PC (in this case, Windows2000 or XP most likely) per the manufacturer's instructions.
Access the Control Panel via the Start menu.
Access network and Internet connections via the Control Panel.
Select the Local Area Connection for the hard-cabled RJ-45 network connection.
Select Properties for the connection.
Set the IP address and subnet mask. Usually you would use a virtual IP such as the following:
IP address: 192.168.127.10
Subnet mask: 255.255.255.0
Check the network configuration and connectivity to the reader.
Select Run from the Start menu.
Type cmd and select OK.
Type ipconfig /all at the command line and press Enter on the keyboard.
Verify the correct IP address and subnet mask in the output.
Check connectivity to the reader with the pingcommand. Type ping 192.168.127.254 and press Enter on the keyboard. Verify reply from 192.168.127.254 in the output.
The reader is now properly configured and can likely communicate to the network. At this point, you can open up either your deployment software such as EasyReader, which can be started on most major reader manufacturers, or you can start the manufacturer's built-in configuration tool.
Figure 7.5: A Properly Connected RFID Reader
The first configuration step is the grandaddy of them all: setting the power level. The best way to do this is to get the reader set up and configured properly in its rack and make sure you are set up with either a laptop or host computer to talk to and control the reader. You should also take a product that is indicative of the products going through the portal and make sure it has an average tag on it (see Chapter 4, "Tags" for how to select an average tag).
Real World Scenario-Want a Reader with That Fry
Back in the early days of RFID, there was a reader called the Mercury II that did not have any sort of power-overload protection. Back then, readers were several thousand dollars each. At the time, I was hiring a new engineer about every month.
It was no surprise that each month we'd get one of these Mercury II readers that just stopped working. It turned out that someone was either powering them up with no antennas attached or was changing antennas with the power on. It didn't take long for us to learn not to power up without antennas on, or for the reader company to add power-overload protection to the reader or at least a 50-ohm terminator to screw on the port when the reader was shipped. There is also the concept of ESD which can fry a reader or tag as well. Electrostatic potential is the ability of a stored charge to essentially zap an electric charge from one point to another based on electrostatic discharge (ESD).
At this point, set the product in the middle of the IZ and start up the reader at the lowest power level. Slowly increase the reader's power until you are able to get a consistent read from the tag. You may want to move the product to several locations in the IZ where you would like to receive the signal from the tag. Make sure you try different box orientations as well, if the boxes are likely to come through the IZ in various directions.
After you have established a lower limit of the read rate, you can look for cross talk and ghost reads. Cross talk occurs when a reader in a dock door that is not receivng any items picks up reads from adjacent or nearby doors. The best way to limit cross talk is to take that same "average" product case and put it on the outer edge of the IZ, where you want to make sure it wouldn't be read. Then, turn up the power until you find that you are able to read the tag even at that outer edge. This will establish your critical boundary.
The critical boundary is the outer edge of the IZ, where you do not want to read any tags. This is important to determine. In fact, it's critical, because you may have a forklift with tag items driving by, or a printer printing RFID tags, or something similar, and you wouldn't want to pick up those stray tags.
After the power levels have been properly set, you are 50 percent of the way to a properly configured RFID reader. The next most common configuration parameter is the antenna configuration. The antenna configuration changes how the antennas each sequence. For instance, if you are at a dock door with four antennas, you can change the way that each antenna cycles in the read order; in other words, if you number the antennas 1, 2, 3, and 4, you can have the reader poll twice on one side of the dock door, so the reader would broadcast from antennas 1 and 2, and then 1 and 2 again before going to 3 and 4. You want to use this antenna configuration to make up for issues that you may have found in your site survey. If one side of the IZ has some equipment that you know is going to reduce the overall effectiveness, you may want to cycle twice on the "weak" side for every time you cycle on the "strong" side.
There are some systems still being used that require reading Generation 1 of the EPC protocol and Generation 2 that necessitates the reader be set to poll both of those protocols. The industry has come a long way over the past couple of years when it comes to creating one interoperable system, so hopefully this will not be an issue in the future.
Unfortunately, the reader companies vary so widely in their configuration processes and protocols that it's impossible to go through each one. There are simple readers such as Conveyor that allow you to choose a configuration and then walk away from a setup system. There are other extremes-for example, a reader such as the Symbol XR series, which has 120 configuration parameters.
Fortunately, there are programs, such as ODIN Technologies' EasyReader, that provide a drag-and-drop interface to configure and tune the readers without having to learn every one of those 120 commands. EasyReader allows you to use a patent-pending portal probe to talk to the reader and automatically tune and configure a reader, and to create a critical boundary and develop real-world RF coverage maps. This software tool makes precise tuning and 100 percent read rates a snap. Figure 7.6 shows an RF coverage map within a designated critical boundary.
Figure 7.6: Critical boundary in EasyReader
The final step of getting a portal completely installed and tested is to make sure that the portal is protected from operator error. The biggest culprit of damaged portals is the forklift driver who has slightly poor judgment. Whacking a reader with a forklift blade, bashing it with a skid, or grinding it with a wheel is a sure way to ruin an expensive investment. The easy solution to protecting the fully deployed RFID portal is to set up a bollard that can withstand a bump or bash from a forklift or other moving vehicle. Figure 7.7 shows a double bollard setup providing ultimate protection of the black portal racks behind the bollards. Some will use a single bollard in front of the rack, but a forklift driver taking a sharp turn can take out the edge of an RFID rack. If I can, I always try to install two bollards at the front of an interrogation zone.
Figure 7.7: Protective bollards in front of an RFID portal
It's ideal if you can install the bollards before you start the testing and tuning because of their metallic nature. You may have to adjust power, angle the antennas, or compensate for any change in the RF field due to the bollards.
Real World Scenario-Reading at the In-bound Doors
One thing that works well if you are trying to read items coming in from trucks at a dock door is to slightly angle the antennas toward the inside of the truck, at a 15- to 20-degree angle. It starts the signal propagating a little ahead of when the tag gets in the center of the interrogation zone to get a response. This is why we made our RFID racks at ODIN with swiveling antenna mounts, to make sure you could angle the antennas to shape the field.
After you have the reader set up, configured, and tuned properly, it's time to prove it. The best thing you can do for a client relationship is to set up the metrics in advance. Let's say they want to read 20 cases on a pallet at an average rate of 98 percent success rate. Get that acceptance testing protocol (ATP) determined before you get on-site and start work. Decide how many tests you will perform and what happens if you are or are not successful.
If your ATP requires you to read 20 tags, make sure those tags are tested first. For information on testing tags, refer to Chapter 4. The tags need to be "average" performing tags, not too sensitive or poor receptors. You want average tags that will represent the typical quality likely to go through the IZ. Acceptance testing and configuration can also be refined after several months of collecting statistics and data about how well each portal is performing. Sometimes, particularly if there are items of different composition being tagged, the readers will need to be adjusted slightly for optimal results.
Real World Scenario-It's All in the Numbers
One of the largest computer manufacturers in the United States hired ODIN Technologies to optimize their RFID network by using our testing tools and scientific methods. Given their more than 250 read points and massive volumes, getting the client from a 98 percent read rate to over a 99 percent read rate could mean millions of dollars for them in return. Following well-proven methodologies, our team was able to optimize the system and get great results for our clients based on the metrics we all agreed on before even starting work.
Several training companies, such as RFID4U and OTA Training, have RFID training classes to prep for the test. In addition to prepping for the test, you'll learn many of the basics of the RFID systems. One of the best parts of these courses is the actual hands-on training. There will be a time when you can play with the readers, learn how to configure them, try to talk with them, and test out some of the software packages that work with the readers. Clearly, the more often you have a chance to use the readers and set them up in various environments, the more you'll be comfortable with how to configure, tune, and test the final IZ. The folks at CompTIA are a great resource for making recommendations for books and training courses. And don't forget their website: www.comptia.org.