Researchers at the Robotics Institute of Carnegie Mellon University in the United States have developed an innovative agricultural sensing technology called SonicBoom, capable of detecting crop positions through sound. This research, published in IEEE Robotics and Automation Letters, provides a new solution for agricultural robots operating in complex environments.

The team, led by PhD student Moonyoung Lee from the Robotics Institute, designed an acoustic sensing system based on contact microphones. Lee stated: “This technology requires no visual input and can determine the 3D position of an object solely through contact.” Experiments showed that the system achieves positioning accuracy ranging from 0.43 to 2.2 centimeters.

Compared to traditional visual or tactile sensors, SonicBoom offers advantages of low cost and high durability. The research team installed six contact microphones inside a PVC pipe, enabling precise positioning by analyzing vibration differences upon contact with an object. Lee noted: “This design is particularly suitable for orchards with dense foliage.”

Currently, the technology can identify hard objects, and the researchers are developing a version capable of recognizing soft agricultural crops. The team is also exploring its potential applications in safety protection for agricultural robots and human-robot interaction. The research is supported by institutions including the U.S. National Science Foundation.