Misplaced items can waste time and, potentially, billions of dollars for major retailers. MIT’s new drone-based system can help.

When you’re dealing with the kind of giant warehouses required by retail giants and other large organizations, taking inventory of stock by hand can be an enormously time-consuming job that verges on the impossible. For example, even the smallest Walmart warehouse is larger than 17 football fields, making it easy for things to get lost. (This is actually more of a serious problem than you might think: over an eight-year period, the U.S. Army lost track of $5.8 billion of supplies in its warehouses.) As a result, companies have increasingly been looking into using drones to speed up the task. However, most attempts to do this haven’t been as efficient as they could be, primarily because they rely on barcode readers or cameras, which miss any items not visible to a camera through line of sight.

That’s an issue that Massachusetts Institute of Technology researchers have tried to address with a new project called RFly, which uses a combination of drones and RFID (radio frequency identifier) tags to, they hope, revolutionize both inventory management and the “non-line-of-sight” problem.

“We developed RFly, a new technology that allows drones to find missing and hidden objects using wireless signals,” Fadel Adib, whose group at the MIT Media Lab developed the new system, told Digital Trends. “Our technology works by analyzing the wireless signals reflected from battery-free RFID stickers. RFIDs are wireless stickers that are attached to objects similar to barcodes. To locate these RFIDs, our drones transmit wireless signals to power them up, then analyze their responses. As these drones fly, they analyze the physical waves of the RFID responses and use these waves to locate the RFIDs. Our technology allows drones to pin down the location of an RFID to the exact shelf an item is on, and our location-finding algorithm is inspired by how airplane radars map the surface of the Earth.”

According to its creators, the RFly system can read RFID tags from more than 50 feet away and identify objects on shelves within 8 inches of their location. The system could also be used for doing more than just Identifying products.

“The applications are vast, and they range from doing remote inventory control in an entire warehouse to allowing people to find missing items at home,” Adib continued. “Imagine a future where each of us has a small miniature drone, and we dispatch the drone to fetch our keys, wallets, or glasses when we can’t find them.”