Demonstration of Natural and Safe Human-Robot Interaction through Speech Recognition and Interactive Robot Skin at ICFP 2021

Author: Zakka Gbouna Vincent

It was a great experience attending the 10th International Conference on Fluid Power Transmission and Control (ICFP 2021) as one of the participants from Zhejiang University. The conference brought together under a constructive platform scientists, companies and engineers throughout the world to present and share their recent research and innovative ideas. Our research work on “Speech Recognition and Interactive Robot Skin for Natural and Safer Human-Robot Interaction” was featured which attracted the attention of the participants with many praising the novelty of the work. The demonstration presented during the conference focused on the intelligent and safe human-robot interaction which was implemented on custom-designed dual-arm robot.

The first function implemented was speech recognition, semantic analysis, object recognition, and grasping detection for natural user-interactive grasping task (grasping an apple, banana or orange on the table and hand it to the conference participant according to his/her voice command).

The second function was the application of robot skin modules with proximity-sensing function for safer human-robot interaction (robot arm will slow down or even stop its operation once detecting the existence of human around, and resume working once human or subject leaves robot’s operation range). It was a great opportunity for us to demonstrate to the participants all around the world the results of our research and also learning from other participants their latest technologies.

The custom-designed dual-arm robot integrated with the robot skin

Interaction with the robot showing visual feedback function of the robot skin