Dr. Swadhin Pradhan and Professor Lili Qiu

PhD graduate Dr. Swadhin Pradhan and his advisor Professor Lili Qiu

When was the last time  you changed out a battery in your house? Studies show that battery-powered devices are used substantially in modern times, Americans use nearly 3 billion batteries every year. Batteries are currently popular because they are able to make electric devices, such as flashlights and watches, portable. However, our usage of batteries extends further from just portable electronic devices. We also use batteries for essential, non-portable electronic devices, such as thermostats and many other sensory appliances. When it comes to such devices, the use of batteries is unnecessary and becomes a hindrance. These batteries must be constantly replaced and over time become time-consuming and costly. A challenge today in the tech industry is figuring out ways to reduce the reliance on batteries for sensory appliances.

One of the main reasons battery devices are popular among sensory appliances is convenience: no one wants a wire stretching from an outlet to a thermostat. To tackle such a problem we need to figure out ways to transfer energy from a consistent power source, such as outlets, without the use of wires. It is this type of research that UT Computer Science researcher and recent PhD graduate Dr. Swadhin Pradhan, with assistance from his advisor Professor Lili Qiu, has dedicated his work to. 

Swadhin Pradhan currently works at Cisco Meraki and researches wireless systems, machine learning, and internet-of-things (IoT) systems. In his research article, “RTSense: RFID based Temperature Sensing,” which was recently presented in ACM SenSys 2020, Dr. Pradhan and Prof. Lili Qiu explores how passive RFID tags could solve this problem of battery-dependent sensory devices. 

Passive RFID tags have existed for almost twenty years now, mainly used for identification and tracking. Passive RFID tags are made up of two components: an antenna and an integrated circuit chip (the actual tag). The circuit chips have no energy of their own—they don’t have a battery and are also not connected to an outlet. They operate purely by absorbing radio frequency energy transmitted from RFID antennas. The radio frequency wave from the antenna energizes the tag, then it sends back information encoded in its memory back to the antenna. These aspects of passive RFID tags make them perfect for sensory appliances. 

The idea behind RTSENSE, the name of the RFID system that Pradhan and his team use to measure temperatures, is fairly simple to understand. Pradhan and Prof. Lili Qiu essentially designed an analytical model that observes the change in electrical resistance in a passive tag caused by temperature change. RTSENSE consists of 2 stages, calibration and estimation. The calibration step is a quick but essential step performed during installation that does not need to be repeated even when tag orientation or distance changes. During this step, the reader continuously reads the tag-pair and records all responses This stage aims to figure out the effect temperature has on this system. With this information, the system is able to effectively estimate temperature in the estimation stage. Temperature is measured by recording the phase difference during 10 seconds. This average is then put in a regression model that estimates temperature. 

Dr. Pradhan measures the readiness of this technology by using three categories to evaluate how well the system performs: a passive RFID tag sensory system must be able to consistently achieve high accuracy, maintain robustness against environmental change, and possess a good sensing range. Previous systems have performed well in an individual category but have failed to perform well in the other two. However, RTSENSE is able to perform well in all these 3 categories. 

Though temperature sensing using passive RFID tags is still not ready to be used commercially, this research marks a significant step in that journey. The potential for such technology is immense. 

Pradhan foresees that such a type of technology will develop enough to where it could lead to personalized temperature control. He envisions a future where this technology allows each room in a house to have the ability to have its own temperature. The commercial implications are also immense—such technology could save companies a considerable amount of money and reduce battery waste. Though this technology is not quite there yet, it has the potential to be, as Pradhan puts it, "a technology that becomes so good, people forget that it exists."This research plays a big role in the journey of this technology to reach that point, it has proven that sensing can be done with passive RFID tags in an efficient manner.

News categories: