Will robots soon see through walls?

An innovation that gives robots the ability to "see through walls"!

A team of researchers at the University of Pennsylvania has developed a new system, called PanoRadar, that uses radio waves supported by artificial intelligence technologies to provide robots with "super vision".

The system works to convert traditional radio waves into accurate 3D images, allowing robots to interact with complex and difficult environments, such as closed spaces, thick smoke or dust, where traditional sensors, such as cameras or "Lidar" (a technology that uses light in the form of a laser to measure distances and detect details in the surrounding environment), fail.

PanoRadar relies on technology similar to beacons that rotate to determine locations, as the device scans the surrounding environment by emitting radio waves, then receiving their reflections to form an image. However, the system outperforms traditional techniques by using AI algorithms that enhance these measurements and generate highly accurate 3D images, similar to those provided by lidar technologies, but at a much lower cost.

With this technology, robots can “see” through physical barriers such as walls or glass, and interact more accurately with their surroundings in difficult conditions such as smoke or fog, where optical sensors are unable to provide accurate information.

One of the biggest challenges the team faced was ensuring accurate measurements as the robot moved, as the system needed to collect data from multiple angles with sub-millimeter accuracy, which required extreme control during movement.

In addition, the researchers trained the system using AI techniques to interpret complex environments, enabling it to identify patterns in radar signals and extract accurate information even in situations where traditional sensors struggle.

In the future, the team intends to expand PanoRadar’s applications to include integration of this technology with other sensor systems, with the aim of developing multi-modal perception systems for robots that enhance their ability to deal with diverse and complex environments, whether in search and rescue operations or in autonomous vehicles.

Source: interesting engineering - https://ar.rt.com/yppa