Internet of Things (IoT) devices are being deployed throughout the world at an accelerating pace, but they have a big problem that is limiting their usefulness — energy storage. Current battery technologies can only supply just so much power before they need to be recharged or replaced, and that makes for a lot of maintenance work. In the case of huge sensor networks dispersed over a large geographical area, keeping the devices powered up is a very time-consuming and expensive process.
So in order to collect all of the data we need to build smarter cities, protect the environment, and improve public safety, we will need to develop more efficient and sustainable energy solutions for IoT devices. One solution that has been proposed in recent years involves removing batteries from the equation entirely with RFID technology. While RFID tags are typically used to identify the items they are attached to, it has been shown that they can also double as sensors by measuring the way that radio waves are changed by environmental factors after being reflected.
Sensor readings are being displayed in an AR view in real-time (📷: UC San Diego)
However, these methods are not widely used because they lack the accuracy of traditional sensors. Accuracy can be improved by measuring the response of multiple RFID sensors simultaneously, but this causes the signals to get jumbled together. But now, a team of researchers at UC San Diego and Qualcomm Institute have described a technique called SenSync that enables them to untangle these signals and capture accurate sensor readings.
Unlike previous attempts to use RFID for sensing, SenSync can process sensor data in real time and with high accuracy levels, even when the RFID tags interfere with each other. Traditional RFID systems were never meant for sensing applications. Their protocols introduce timing mismatches and signal distortions that make precise measurement nearly impossible. But SenSync solves these problems using a modified version of dynamic time warping, a signal alignment technique originally developed for speech recognition.
This algorithm allows SenSync to take in slightly misaligned or out-of-sync data streams from multiple passive RFID tags and combine them into a coherent, accurate signal. As a result, the system can deliver up to five times higher resolution and eight times the data throughput of previous approaches. That is up to 500 samples per second with sub-degree error rates, even in cluttered or dynamic environments.
Data packets from multiple RFID tags are aligned (📷: I. Bansal et al.)
Using SenSync, passive RFID tags can reliably monitor things like temperature, pressure, moisture, and weight, all without requiring any batteries, wires, or other custom electronics. That opens the door for ultra-low-power sensing in everything from smart agriculture and medical monitoring to warehouse automation and AR-enhanced interfaces.
Ultimately, the researchers believe that SenSync could be an important step toward what they call physical AI, or systems that sense the real world as easily as today’s AI models process text or images. With trillions of RFID tags already in use worldwide, this work could transform the IoT landscape into something far smarter and more connected.