As per the statement published in the journal Nature Communications, a team of scientists at the University of Washington has devised a solution where noise-canceling robots can block noisy conversations in crowded places like cafes and restaurants.
Intelligent sound robots have been developed to function as speakers that automatically disperse in the room, dividing large rooms into acoustically isolated zones and tracking the locations of people speaking.
This type of robots benefit from sound technology similar to headphones that cancels noise. These small devices placed near people speaking loudly can emit sound waves that counteract their sound waves, effectively canceling the noise produced by their conversations.
The project is still in its early development stages, comprising a “swarm” of robots using various techniques such as self-motion, sound analysis, and voice recognition. Advanced deep learning algorithms are prominent features aiding the robots in detecting high-volume sound locations, distinguishing distinctive voices, and recognizing speakers.
Scientists envision using these robots when conversations in a specific group at a cafe become loud, where the robot near the group can reduce conversation noise without interfering or disturbing neighboring speakers. Additionally, robots can replace the central microphone in meeting rooms, providing enhanced sound control in those spaces.
Doctoral student at the College of Engineering and Computer Science at the University of Washington and project participant, Malik Aitani, stated: “If I close my eyes and there are 10 people talking in the room I am in, I won’t be able to discern who is saying what and where their voices are coming from in the room. It has been extremely challenging for the human mind and also for technologies until now. But for the first time, we can now track the locations of multiple people speaking in one room and separate their conversations using what we call the audio robot swarm.”
The swarm consisting of 7 robotic microphones, each with about a one-inch diameter, connected to a central charging station. The robots extend and retract autonomously to and from the charging station, allowing them to move easily within the area. By generating high-frequency sounds and studying their reflections in the surroundings, the microphones can determine their location and autonomously direct themselves towards different tables (similar to bats). This capability enables them to map the area, avoid obstacles such as chairs, tables, and people, effectively contributing to segregate and organize noisy areas.
Tushar Chin, a project participant, explained that if there is one microphone at a foot’s distance and another two feet away, the sound will be first received from the microphone that is one foot away. And if there is another person closer to the microphone that is two feet away from me, their sound will be first received from there. We have developed networks using these delayed waves to differentiate what each person is saying and track their location. Thus, conversations can be conducted between four people with isolation of each speaker’s voice and determining each person’s location in the room.
Therefore, the robotic system continuously monitors noise levels and organizes the robots’ locations and activities to effectively reduce noise in all rooms. This system has been tested in various rooms, including workspaces, living rooms, and kitchens, with groups of three to five individuals speaking. In one of these tests, the system was able to differentiate between different sounds even when the individuals were 50 cm apart.
According to the team of scientists, they are also exploring the possibility of developing robots capable of emitting sounds that prevent their spread in specific areas while allowing it in other areas, with the goal of enabling individuals in different parts of the room to hear diverse sounds.
