SonicSense Revolutionizes Robotic Object Perception with Acoustic Vibrations

Duke University's new system enables robots to identify materials and shapes by 'hearing' vibrations, enhancing their interaction with the physical world.

• SonicSense uses contact microphones in robotic fingertips to detect vibrations from objects, allowing robots to determine material type, shape, and contents.
• The system leverages AI to analyze frequency features from interactions, enabling accurate identification of new and known objects.
• SonicSense performs better than previous methods by utilizing multiple fingers, advanced AI, and noise-filtering techniques.
• This innovation allows robots to operate effectively in dynamic, unstructured environments, bridging the gap between controlled lab settings and real-world applications.
• Future developments aim to integrate additional sensory modalities for more complex interactions, with potential applications in cluttered and dynamic environments.

3 Articles

ScienceDaily Neuroscience News Tech Xplore