Researchers from University of St. Andrews have created a new system called RadarCat that can recognize virtually any object simply by touching it. Without the use of any cameras or visual sensors of any sort, RadarCat quickly deduces whatever object is placed on it, distinguishing between apples and oranges, smartphones and tablets or anything else.
RadarCat uses Google's Project Soli as its basis. The Google project, which was completed last year, is an experimental smartwatch sensor that can recognize even the tiniest of gestures. Though it was never made available commercially, Google shared the tech with select labs, including the researchers at St. Andrews.
RadarCat is so sensitive that it can recognize whether a glass is empty or full, and it can even recognize what type of liquid is inside. As such, one of RadarCat's potentially limitless use cases could be to automatically alert servers when consumers need a new drink.
RadarCat Uses Google's Project Soli to Accurately Predict Any Object
References: sachi.cs.st-andrews.ac.uk & fastcodesign