In warehouses, factories, and other dense or crowded settings, finding specific items can be tricky, especially if they’re hidden by the many other items. Even conventional robots aren’t much help if the target item is obscured from the robot’s camera.
Tara Boroushaki is here to help.
As a graduate student at MIT, she developed FuseBot, a robot that uses radio frequency to quickly root out hard-to-find items even when they aren’t in plain view. More generally, she develops algorithms and builds robotics and other systems that can interact with the environment more efficiently and capably.
In the latest edition of “Office Hours,” a Q&A series that introduces new Yale faculty members to the broader community, the assistant professor of electrical and computer engineering discusses her research, the connection between her work and magic, and banana bread.
| Title | Assistant professor of electrical & computer engineering |
|---|---|
| Research Interest | Sensing and mobile technologies with applications in wireless networking, wireless sensing, cyber-physical systems (including robotics), and cyber-human systems. |
| Prior Institution | MIT |
| Start Date | April 2025 |
How would you describe your work?
Tara Boroushaki: The majority of my research is about how we can use different types of signals alongside other modalities — like computer vision — to sense the environment in a new way. For example, what is good about signals like WiFi and Bluetooth is that their signal can go through walls and go through obstacles. That’s why you can get WiFi when the router is in a different room.
In warehousing logistics, when items are missing or lost it is a source of wasted money. Being able to find these items, even if they’re not in the line of sight, is a very important task. With our robot, we’ve shown that we are at least twice as fast as previous work, and sometimes more. When you scale that to a huge operation, like an Amazon warehouse, the amount of time and money you save is pretty high.
What makes your system so fast?
Boroushaki: Well, if you’re trying to find a specific item in a pile of returns in an Amazon warehouse, a normal robot with only cameras would need to explore the whole pile. We showed that if we only use the information from even a couple of items that are tagged with radio frequency identification [RFID], the robot can reason about how much space each of those items will occupy in the structure of the pile. So if I know there’s an iPhone here and this space is the size of an iPhone box, then I know a tablet can’t also be here, because physically that’s impossible. We’re creating this type of reasoning for a robot to be able to understand that. On average, that makes the system twice as efficient.
How did you get interested in this particular field?
Boroushaki: I was interested in being able to see the hidden world. Basically, I wanted to do something that I cannot do with my own eyes or my own senses — I was interested in sensing what my eyes cannot see. That’s why I got interested in wireless signals, which go through obstacles and come back.
Growing up, were you always interested in science and engineering?
Boroushaki: I was more interested in magic. Yeah, I was waiting for my letter from Hogwarts. I didn’t get it. Then I had to go to MIT.
Did you see what you do as the closest thing you can get to magic?
Boroushaki: Yeah, honestly, I think so. I think I was interested in super powers and magical traits and stuff like that. I think wireless sensing and perception in non-line-of-sight is pretty close.
Speaking of Hogwarts, do you have a favorite character in Harry Potter?
Boroushaki: Hermione Granger. That was one of the first questions my advisor asked.
What do you like to do when you’re not in the lab?
Boroushaki: I do a lot of baking — anything that I find on the internet. I do a lot of desserts, like banana bread. There is this room at MIT, called the Banana Lounge, that has free bananas every day. That was great because you could always find bananas ready to turn into banana bread.
