Researchers at MIT have developed a robot that uses radio waves to sense occluded objects to provide a better way to grasp items that might otherwise be blocked from view.
Called RF-Grasp, the robot combines sensing with more traditional computer vision to locate the hidden objects and then grasp them to be moved to another location. MIT said this system could one day be streamlined into an e-commerce fulfillment warehouse.
“Researchers have been giving robots human-like perception,” said Fadel Adib, an associate professor at MIT. “We’re trying to give robots superhuman perception.”
While industrial automation has become a staple in warehouses, humans still play a large role despite sometimes-dangerous working conditions. Robots in warehouses struggle to locate and grasp objects in such a crowded environment, MIT said. Optical vision helps but robots cannot perceive the presence of an item packed away in a box or hidden behind another object on a shelf.
How it works
That is where radio waves come in. Radio frequency identification (RFID) is used to track a variety of objects from library books to pets to packages to pallets. In an RFID system, the tag is a tiny computer chip that gets attached to or implemented into an item that is tracked. The reader then emits an RF signal that is modulated by the tag and reflected to the reader.
This signal gives information about the object itself as well as its location. MIT believes RFID could be a boon for robots, giving these machines a mode of perception.
RF Grasp uses both a camera and an RF reader to find and grab tagged objects, even if the object is not in view of the robot’s camera. The system includes a robotic arm attached to a grasping hand, which features a camera attached to the wrist. The RF reader is independent of the robot and relays the tracking information to the robot’s control algorithm, allowing the robot to constantly collect both RF tracking data and a visual picture of its surroundings.
“The robot has to decide, at each point in time, which of these streams is more important to think about,” said Tara Boroushaki, a research assistant at MIT. “It’s not just eye-hand coordination, it’s RF-eye-hand coordination. So, the problem gets very complicated.”
RF Grasp can zero in on a target object and as it gets closer the robot’s decision making is based on the vision and RF details it receives.
MIT said compared to a similar robot equipped with only a camera, RF Grasp can pinpoint and grab a target object with about half as much total movement. The robot can also declutter an environment by removing packing materials or other objects in order to find the targeted object.
The robot could one day perform fulfilment in packed e-commerce warehouses and could instantly verify an item’s identity without the need to manipulate the item, expose its barcode and then scan it.
RF Grasp could also potentially be used in home applications such as for finding tools, keys or other objects that are tagged but lost.