Imagine a keyboard that senses not only which key you pressed but also with how much force. American researchers from Washington University in St. Louis (WashU) have taken a step toward such technology using just one tiny optical sensor. The work was published in the journal Laser & Photonics Reviews.

Conventional sensors usually measure one thing. To capture several parameters at once, engineers have to install several sensors - the device becomes larger, more complex and requires more labor-intensive setup. The group led by Lan Yang (professor of electrical and systems engineering) and Chenyang Lu (professor of computer science and engineering) took a different path.

The scientists created a miniature device that holds light inside itself - an optical microresonator. Light in it travels along many paths at once. When something changes around it, each light pattern responds in its own way. Together these responses form unique optical "barcodes" carrying information about the changes. By analyzing how such "barcodes" change, one can learn several things at once - for example, exactly where the surface was touched and with how much force it was pressed.

Machine learning - an artificial intelligence method in which a program learns from examples - helps decode the complex light patterns. The algorithms find hidden regularities in the optical signals and turn them into useful measurements.

To test the approach, the engineers assembled a miniature keyboard-like platform with nine touch points. Instead of nine separate sensors, they used one optical microresonator. It determined both the location of the press and the magnitude of the surface displacement. Tests also showed that the method works when optical measurements become noisier or contain less data - this is important for practical systems.

"One tiny optical sensor can perform functions that usually require multiple sensors," Lan Yang was quoted as saying in a university press release. According to her, this opens new possibilities for creating smaller and smarter sensor technologies.

Potential applications go beyond keyboards. They include robotics, where a sense of touch is needed, health monitoring and cybersecurity. For example, a smartphone could recognize not only the entered code but also the nature of the presses - say, the pressure applied to different digits. This would add an extra level of authentication, co-author Hanyue Li suggests.

The authors emphasize that this is so far a laboratory demonstration, not a finished product. Further development lies ahead. But the principle already shows that instead of increasing the number of sensors, one can extract more information from light with the help of artificial intelligence.

Source: Tech Xplore