Carnegie Mellon University
June 07, 2019

Researchers Push Technology To Lessen Lost Luggage Woes

By Olivia Olshevski

Sherry Stokes
Researchers at Carnegie Mellon University are searching for new ways to improve the tracking technology so airline passengers won't have to look for bags at the lost and found counter.

Most airports use a radio-frequency identification (RFID) system to track the movement of your baggage throughout the airport. This system contains two parts: a mobile tag and a stationary reader. While the reader remains at a fixed checkpoint within the airport, the tag is attached to the baggage and serves as a unique label. As a suitcase moves through the airport, it passes certain baggage checkpoints. When the tag comes within range of the reader, the two parts of the system begin to communicate with each other, transmitting and receiving signals. Through this communication, the reader can confirm that a baggage has made it to each checkpoint.

Checkpoints are placed at specific points throughout the airport, often appearing periodically along the path that the baggage must take from check-in to boarding. Unfortunately, each reader has a range of five to 15 meters compared to the distance between checkpoints. In other words, there are blind spots along the path, where no checkpoint can confirm that a bag is nearby. Once baggage is outside of the range, it becomes essentially lost — even though it is most likely still somewhere along the proper path.

Many warehouses, such as those used by Amazon, pair merchandise with RFID tags to track the shipping process, and some retail stores place tags on clothing to keep tabs on inventory. These items are often lost in similar blind spots created by the RFID readers.

Solutions to the problem have been proposed. Some researchers have focused on changing and improving the readers themselves, while others have altered the tags' hardware to achieve better ranges. However, both options require the installation of new RFID infrastructures.

Recently, researchers in the lab of CMU's Swarun Kumar have developed an update for current system software to increase the range.

"From a deployment perspective, upgrading software on existing readers is much more inexpensive compared to purchasing and installing new readers that are often bulkier," said Kumar, an assistant professor in the Department of Electrical and Computer Engineering.

"Our solution, called PushID, uses a technique called beamforming that focuses energy from many different readers on to one tag. By carefully modifying the signals from each reader, we make sure their energy constructively adds up at the tag's location," Kumar said. "Our key innovation is finding where the battery-free tags are to beam energy to, because they have absolutely no energy in the first place to advertise their location."

To determine where tags are located throughout the environment, the readers give out various specialized signals that intelligently smear energy through the environment in search of a tag. If a tag is within the range, it will transmit a signal in return. The reader will then receive this transmitted signal and once again sends out a signal of its own. By repeating this process every few milliseconds, the readers can quickly identify the tags in the environment and can converge on their precise locations.

This iterative process also allows PushID to account for any obstacles in the environments, such as furniture, signs and walls. The presence of even a single obstacle can dramatically change how energy adds up or cancels out across tag locations. PushID allows the RFID system to account for the effect of these obstacles, thereby ensuring that the tag's location is properly identified. The readers can also account for mobility, allowing the system to follow the tag as it moves through the environment.

Unlike current RFID tags, which have a range of 5-15 meters, PushID can expand that range to 64 meters. And for a given environment, PushID can provide sufficient coverage for 97% of the area within four seconds; without PushID, only 33% of the total area is covered.

While PushID is ready for implementation as it stands, the team has already identified additional improvements. The range of the PushID system can be expanded if the team is allowed to choose where new readers should be installed. Additionally, the team hopes to incorporate their technology with other solutions currently on the market to further increase range.

Currently, PushID works at a store-scale: 64 meters can cover a warehouse or retail store, but it is inadequate for larger areas. In the future, the team aims to bring the technology to a city scale and broaden its applications.

Jingxian Wang, the Ph.D. student in ECE who spearheaded the project, dreams of a world in which PushID has expanded applications.

"The technology may one day allow us to track our phones and our clothes, every item that we don't want to lose, throughout entire cities," Wang said. Far from the days of losing your luggage on important trips and trying not to pull your hair out, PushID may be the answer to never losing anything again.