Dec 27, 2015The Internet of Things (IoT) needs sensors. The ever-increasing desire to measure our environment has already cut out the cables from the sensors by means of such technologies as ZigBee and Bluetooth Smart. However, a sensor node's lifetime is usually limited to battery life. This can be extended using expensive energy-scavenging solutions. However, a lot of deployment sites lack potential energy to be scavenged, and the accumulators used with scavengers tend to have a limited amount of charging phases.
Thinking passive brings us to the world of radio frequency identification. Low-frequency (LF) and high-frequency (HF) RFID solutions hardly bring the benefit of wireless sensing due to short reading distances. They suit applications in which measurement points can be brought within a few inches of an RFID reader. Only ultrahigh-frequency (UHF) RFID can truly remove the cables, but the UHF option brings additional problems to the solution due to the limited amount of energy available from a UHF reader's signal.
Adding a sensor to a UHF EPC Gen 2 passive RFID tag is not a novel idea. Several products are already in the market and new ones are being developed. However, there is a common approach with all current RFID sensors: the idea of using digital communication for sensor-value readout. This approach has several limitations. The sensing element provides analog output, which is changed to digital format via an analog-digital (AD) converter. Sometimes, the value is transmitted between the sensor element and the RFID chip using an I2C or other bus. The value is possibly stored for a while, waiting for the RFID communication slot. All these processes require an excessive amount of energy compared with reading a simple Electronic Product Code (EPC) number encoded to a tag. Moreover, there is a mandatory delay by the protocol when additional memory banks are accessed to get the sensor value out of the tag.
Metso has been conducting wireless research in cooperation with Aalto University's radio science group in Helsinki. The goal has been to create a solution capable of tackling the drawbacks of present-day UHF RFID sensor approaches that would still fit within the common standard. The research is done and the results are enlightening.
UHF RFID communication is typically considered to be digital communication. The EPC is encoded as bits, as are all other commands and responses. The communication is specified in the UHF EPC Gen 2 RFID standard and made very efficient. However, EPC Gen 2 (and the related ISO 18000-63 standard) provides an additional layer of communication that does not affect the digital part. The modulation frequency of the backscatter link is specified with great tolerance (for example, ±10 percent). This is primarily due to RFID chips' poor RC oscillators and sensitivity to temperature changes. If we can have a stable oscillator with a high Q value and eliminate the temperature sensitivity, we can achieve a great bandwidth for transmitting extra information from the tag to the reader. That's exactly what we did. The given ±10 percent variance allowed in backscatter link frequency (BLF) results in 51 kHz of bandwidth with a 256 kHz BLF and 64 kHz of bandwidth with a 320 kHz BLF. Even with safe margins, one can easily fit a single analog sensor value in it.
The new innovation is branded as TuntoID. Since the development is still in its early stage, we have not yet determined all of the benefits that can be acquired. However, at least some are clearly seen. The longer read range is mainly due to the lower power usage. The final readout distance is still unknown, but from the latest measurements, we can see that it will be on par with the best available UHF RFID tags. Faster readout of sensor values comes from the need to read only the Electronic Product Code. This leads to a sampling interval of several milliseconds.
The approach could also support a capacitive, inductive or voltage-generating sensing element, as long it met the requirements for a very low-power analog interface. These kinds of sensing elements are mostly MEMS-based and can be obtained from several providers. The sensing element can be located directly on the tag's surface or further away, providing both a good sensing interface and a better location for the antenna. Multiple backscatter frequencies can be loaded with different sensors. In practice, a single tag can hold three different sensing elements that are read one after another.
Reaching the Unreachable
The boundaries of passive wireless sensing have been pushed a big step further. This approach brings one additional layer in UHF RFID communication. In the future, any EPC Gen 2 RFID tag may also hold an additional sensor that is interrogated every time the tag is accessed, as long as the reader knows what to look for. My colleagues and I are now seeking out applications that are within these new possibilities—applications that have been considered impossible due to the need for long-range or fast sampling.
The hype around the IoT has been around for a few years. Once the hype fades away, we will still have places that current measurement technologies cannot reach. There are the sweet spots for something new—something like TuntoID.
Joona Nikunen is a Metso senior research engineer who concentrates on industrial wireless challenges.