If you have ever used an access card to get into a building or your car or passed through an automated toll collection system on a highway, you have used RFID. The definition of RFID is rather broad because it has so many uses. Let's first start with the acronym and what it means:

RFID = Radio Frequency IDentification

Now, let's break down what this means: a system of technologies that allows an object, person or animal to wirelessly identify itself to another object, person or animal. Hence the words RF (Radio Frequency) and ID (IDentification).

To be able to do this in so many usage scenarios, form factors, price points, thermal environments, et cetera, the technology used for enabling RFID takes many forms. The most common ways of subdividing the technology are by frequency and whether or not the tag is a passive device. Let’s first look at the frequency: at the lowest common frequency or LF (Low Frequency), this spans the range of 58-148.5 kHz or 58-148.5 thousand cycles per second. This frequency allows for low cost tags and readers with short read range (several inches to several feet), but most importantly, this frequency allows the RF to transmit through metals a few mm thick and liquids. This makes this technology very suitable for implanting into animals, but also for access control and electronic article surveillance (EAS) or antitheft applications. Now, one can't easily implant a battery along with the antenna and chip in animals or consumer goods, so the tags are read passively. The behavior of the tag changes an incident RF field in a way that a reader can detect a unique ID. This ID may be a single bit in the case of an EAS tag or up to 10s of bits for animal tags.

The next frequency range spans from 1.75-13.56 MHz and is the next most common use of the technology. This frequency range is called HF, or High Frequency, and includes tags for use in building access, public transportation and electronic payment systems. The range of these systems is similar to LF: inches to feet, depending on the application. HF tags also work pretty well with metals and liquids. Except for electronic article surveillance, HF tags are usually used for proximity applications: a human gesture of moving one's arm, wallet or purse is used to provide access or payment.

Frequency Allocation
Click to Enlarge

Let's briefly touch on the next frequency range: UHF or Ultra High Frequency, sometimes called Near Field UHF. This is the primary area of focus for ThingMagic because we believe UHF allows RFID to be used ubiquitously, that is, where the technology can be used in every place and every time. How is this possible? UHF RFID spans the 433, 840-960 MHz and the 2.4 GHz range. At this frequency, the RF that is produced allows for relatively efficient wave-like propagation; similar to a radio station, but with reasonable amounts of power consumption for handhelds, laptops, trucks, printers, et cetera. Near field UHF RFID tags contain a small silicon chip and an antenna paired onto or into an object. FilesThis allows one to create tags which can be read from inches to 10s of feet in a passive configuration, and 100s to 1000s of feet if used semi-passively (the tag still changes the RF that comes back to the reader rather than transmitting, but gets a battery to help it out) or actively (an active transmitter). The tags can also be produced very inexpensively; the antennas can be etched with chemicals or printed with a printer that can print metals such as copper or aluminum, or in the case of a chip produced by Hitachi, directly into the chip itself.

The low cost and long range of UHF RFID means that tags can be placed just about anywhere and interrogators (or RFID readers) can read them. This allows computers attached to these interrogators to see the world around them; not with the lens of the visual spectrum like humans do, but through the RF lens. It is a pretty primitive way to see the world compared to the visual acuity and processing that human beings are capable of, but this technology can be made to work very well and without human intervention.  It can reduce some hard machine learning problems to simple observational ones. The technology is now starting to be used ubiquitously in hospitals, law offices and courthouses as an ambient computer interface for objects. In the courthouse example, the technology is used to track all documents and docket folders within their environment. Previously, near the time of a hearing, a long and laborious process was manually run with human labor to locate needed paperwork. Now with this new technology, the analog of an internet search can be made to search for objects in the physical world (Ravi Pappu coined this a reality search engine). The search is done in the digital world, but the association with identity and location in the physical world allows one to seamlessly cross this digital-physical barrier. These applications point to a future where identification and sensing could become ubiquitously present in every object in a user environment. This would allow the creation of one of the ultimate ambient interfaces: each application would be created by filtering relevant data for a particular use case.