Korean researchers have developed a vertically stacked semiconductor system that can process both current and past signals, potentially allowing small wearable devices to analyze movement and physiological data without relying heavily on external computers.
The technology was developed by researchers from KAIST, UNIST and POSTECH. It uses carbon nanotube transistors controlled by a solid-state ionogel, a material that contains mobile ions that affect how electric current flows through the device.
When voltage is applied, ions move through the material. They do not immediately return to their original position after the voltage disappears, leaving a temporary electrical “memory” of previous signals. Instead of viewing this delayed response as a limitation, the researchers used it to process information that changes over time.
By adjusting the ion concentration and thickness of the ionogel, the team created devices with different response rates. The researchers then stacked fast- and slow-responding devices vertically so that different layers could process information over different timescales.
Different layers remember differently
The faster layer is more responsive to the latest input, while the slower layer retains information collected over a longer period of time. Therefore, by combining their outputs, the system receives information about what just happened and what happened immediately before.
Previous ionic materials used for such devices were often liquids or soft gels. These materials can make precise semiconductor manufacturing and stacking multiple layers difficult. By converting the ionogel into a solid thin film, the researchers were able to better control its properties when building multiple device layers.
The team tested whether the hardware could distinguish temporal patterns using four consecutive input signals, each of which could be turned on or off. The devices successfully differentiated all 16 possible combinations.
The researchers also used the devices’ measurements to create a simulation to classify videos of moving handwritten digits. When processing moving image sequences played at different speeds, the system achieved validation accuracy of over 90 percent.
The approach could eventually help electronics process changing information locally, rather than constantly transmitting raw data elsewhere for analysis. Potential applications include wearable devices for monitoring movement and physiological signals.
Wearables gain local memory
The team demonstrated that the technology is compatible with various manufacturing formats. The researchers fabricated the devices on 4-inch wafers as well as flexible substrates, potentially making the architecture suitable for electronics that need to conform to non-flat surfaces.
According to the researchers, the devices maintained stable electrical properties even 55 months after manufacture.
“These devices can be fabricated on large-area substrates using existing thin-film semiconductor processes and can also be stacked in multiple layers,” said Professor Jimin Kwon of KAIST.
Kwon said the technology could be further developed into chips that analyze movements and physiological signals in low-power devices such as smartwatches.
However, the researchers have not yet been able to demonstrate these energy savings in a real smartwatch. Further testing is needed to determine how the architecture works and how much energy it can save when integrated into real wearable electronics.
The study was published in the scientific journal Advanced materials.
