Researchers at MIT have developed a robot called RF-Grasp that uses radio waves, which can pass through walls, to sense objects that you cannot see at first.
RF-Grasp combines sensors with traditional computer vision to identify and capture potentially invisible items.
And progress could one day simplify warehouse e-commerce tasks or help a machine snatch a screwdriver from a hybrid toolkit.
Fadel Adib, an assistant professor at the Massachusetts Institute of Technology, says: Researchers were giving robots a human-like perception, and we are trying to give robots a supernatural vision.
As e-commerce continues to grow, warehouse work is still usually the domain of humans, not robots, despite sometimes hazardous working conditions.
This is partly because robots struggle to locate and capture objects in such a crowded environment, as perception and selection are barriers in the industry today.
Using visual vision alone, robots cannot perceive an element hidden in a box or hidden behind another object on the shelf, as visible light waves do not pass through walls, but radio waves can.
RF Grasp uses both a camera and an RF radio frequency reader to find and capture tagged objects, even when the camera is completely hidden from view of the camera.
As long as the item includes the RF tag, the robot can find it, even if it is hidden behind objects, such as wrapping paper.
And unlike typical university robotics projects, there is a clear use case for this invention as well, and the team believes that RF Grasp is helping companies like Amazon increase and simplify their warehouse automation.
The most challenging aspect of developing RF Grasp is integrating visual vision and radio frequencies into the decision-making process, and the team compares the current system to how you react to a distant sound by turning your head to pinpoint its source.
RF Grasp uses an RF radio frequency reader to find marked objects, and the closer it is to an object, the more it relies on the information it collects through the camera for decision-making.
Compared to a robot that operates on the visual system only, the RF Grasp robot can locate and capture the object in about half of the total movements, and it has the unique ability to clean its workspace while it is performing its tasks.
There is a trend in countries, such as Japan, to spread radio frequency tracking tags in the retail sector, which means that the technology is in place for RF Grasp to work.