We have reached a significant milestone with GE’s radio-frequency identification (RFID) sensors! Our GE Global Research team (see Figure 1 below) has developed sensors for detection and quantitation of chemical threats such as explosives and oxidizers and has tested these sensors in collaboration with our partners. The significance of this accomplishment is in accurate quantitation of minute amounts of these chemicals with our individual RFID sensors outside the pristine conditions of GE labs.
We have reached a mid-way point in development and commercialization of these sensors that will complement conventional analytical instruments for detection of chemical threats. In airports today, chemical threats are often screened using desktop systems — suspicious surfaces are swabbed and separately analyzed, consuming substantial time, space and power. Compared to such desktop detectors, our sensor system is 300 times smaller, weighs 100 times less, and uses 100 times less power. Also, compared to the arrays of multiple sensors needed with desktop detectors, our RFID sensors so sensitive that we achieve accuracy by using only individual sensors.
Of course, the question is – how can a single simple sensor compete with the detection performance of more sophisticated conventional analytical instruments or sensor arrays? Indeed, our sensors look very simple (see Figure 2 below), but there are four key new features that enable their desired performance – a sensing material, a matching transducer, a sensor reader, and data analytics. Together, these features make our sensors “multivariable sensors” and boost their performance without a change in their appearance. As well put by William Shakespeare, “appearance can be deceiving.”
To build our multivariable RFID sensors, we carefully design a sensing material for a particular application scenario and match its response with the right geometry of our RFID sensor antenna. The multivariable sensor response is measured using a cell phone-sized sensor reader device. The sensor reader is responsible for the accuracy of sensor response and its ability to correct for fluctuations of ambient temperature and other environmental instabilities. A more conventional way to measure responses of RFID sensors is to use near-field communication (NFC) that is available in many modern smartphones. At present, commercially available RFID sensors and NFC phones are successfully applied for quantitation of humidity and temperature (Read more here, here, and here.) NFC sensors will continue to expand their applications in situations where sensor response is not expected to suffer from interferences. Otherwise, NFC sensors cannot provide detection selectivity and require conventional arrays of sensors with their well-known practical challenges. Our individual RFID sensors solve this problem by having all data analytics “smarts” located in the sensor reader, rather than in the sensor. Apple founder Steve Jobs once said that “software is going to be a major enabler in our society.” In our sensor applications, we see data analytics as the key enabler in achieving sensor performance.
The principles of our data analytics are shared among different types of multivariable sensors that we are developing at GE. Indeed, while our multivariable sensors can be based on RFID or bio-inspired sensing principles, the common themes include collection of response from a simple, cost-effective, and often single-use multivariable sensor and data processing using a non-disposable sensor reader so all the “smarts” reside in the sensor reader.
Our GE teams are developing multivariable sensors to perform measurements with accuracy and reliability in complex environments, confined spaces, and without available external power.
But what if in future it isn’t just sensors that are inexpensive – less than 50 cents each – but also the sensor readers? What if the readers cost less than a cup of coffee?
Stay tuned… Share your thoughts… This future is here!