The innovative device uses a unique approach known as: “robotic acoustic swarm”. The speaker consists of seven microphones, each the size of a thimble, that communicate and maneuver independently like miniature Roombas, moving to optimize sound capture and returning to the charging dock when necessary.

said the study’s lead author, Malek Itani, from the Paul H. Allen School of Computer Science and Engineering.

For the first time, we were able to track the location of multiple people talking in a room and separate their speech using what we call robotic “acoustic swarming.”

The novelty of this technology lies in the mobility of robots equipped with microphones that use the high-frequency echolocation technique to navigate the environment. Thanks to the strategic distribution of microphones, the neural network responsible for processing the data can make extremely precise calculations.

However, it is important to remember that robots are currently limited to their functions in two-dimensional space, moving mostly on the tabletop.

Mobile microphones can move independently on the desktop / Photo: Nature Communications

Tuochao Chen (Tuochao Chen), co-author of the study, spoke in detail about the functionality of the system and explained:

“We developed neural networks that use these time-delayed signals to separate what each person is saying and track their location in space. So you can have four people have two conversations and isolate any of the four voices and locate each of them in the room.”

Speaker system based on an array of robot microphones: watch the video

The potential applications of this revolutionary technology are vast, and one of the most promising areas is video conferencing and meetings, where the ability to shut out distracting or unwanted noise can greatly improve communication and productivity. Additionally, a speaker system’s ability to isolate simultaneous speech may be important for improving voice recognition and individual voice perception.