Google #Project Soli, the tech giant’s venture into gesture technology, has developed the capability of sensing objects and differentiating between them. In the making since 2015, the project uses CatRadar to accurately pinpoint the position of an object and its relative motion. Now, a Scottish research team from University of St Andrews lead by Professor Aaron Quigley has uncovered a new facet of the nascent technology: object detection.

Radar Categorization for Input and Interaction or CatRadar is a very promising material and motion detection technology. Researchers in Scotland exploited #Google Project Soli developers kits to teach Google’s recognizing objects using a miniature radar.

Advertisements
Advertisements

This opens up a new potential of Soli's applications which originally intended for gesture detection, not objects.

"Radars have been used for many different things, to track cars, big objects, satellites, and planes; we are using them to track micro-motions," Project Soli founder Ivan Poupyrev said in the introductory YouTube video (below) by Google ATAP. With the new findings, Soli's CatRadar has turned from a control-adjusting input device to an object detector by utilizing unique radar fingerprints, a natural property that exists in tangible objects.

Benefits of CatRadar

  • CatRadar’s miniature size enables them to virtually fit in anything including smartwatches.
  • The radar technology can track movement to pinpoint accuracy and can read a motion of sub-millimeter proportions.
  • Unlike Internet of Things (IoTs), no WiFi connection would be required to send inputs across various devices.

Uses of Project Soli

Once Google Project Soli is rolled out fully-furbished, possibilities are endless.

Advertisements

That’s because developers can set any type of motion to control a specific parameter of a connected device. For example, users could control WiFi Speakers by simply turning an imaginary knob in the air. This gesture could alternatively be sliding the thumb across index finger to act as a virtual slider in an app or a game. Some of the current capabilities of CatRadar is shown in the video below.

Future perspective

The miniature radar system is so accurate it can differentiate between "the front and back of a smartphone, or tell whether a glass is full or empty." Projecting further applications of this property, it could act as a highly accurate gauge to measure the volume of fluids (liquid fuels etc) in a container down to milliliter accuracy. Likewise, it’s material distinguishing properties could revolutionize automated security check to a point where metal detectors would seem primitive.

Project Soli challenges

The Scottish research team also unveiled some of the challenges faced by Project Soli if it pursues the path of sensing objects.

Advertisements

CatRadars are incredibly accurate but they would need some logic in the form of algorithms. Quigley told The Verge about the limitations that it "does occasionally confuse objects with similar material properties."

The head of the research team also pointed out that Google Project Soli's CatRadar also faces difficulty detecting hollow or oddly shaped objects, which makes sense because the amount of information fed back to the sensors depends on up the shape of the object. Objects with flat surface bound back more information than those complicated shapes. #News