Shared Control Architecture Based on RFID to Control a Robot Arm Using a Spontaneous BMI

This paper describes a shared control architecture combining a Brain-Machine Interface (BMI) with Radiofrequency Identi?cation (RFID) technology. RFID information is used to solve the limitations of the BMI, which is only capable of generating up to three different commands. A real application has been designed consisting of a setup with an object that can be picked and placed by a robot arm. Users control the robot arm generating three mental tasks related to motor imagery. To detect the mental tasks, 16 electrodes have been used to register the brain activity. The feature extraction algorithm is based on a combination of Fast Fourier Transform (FFT) and Wavelet Transform (WT), while the classi?er is based on Linear Discriminant Analysis (LDA). Four volunteers have controlled the robot arm to perform a particular pick and place operation, and time and accuracy have been measured. The results show that users are able to move and place objects on their own will, using only three different mental states, thanks to the shared control architecture.

Category:

Science & Technology

Tags:

• Brain-Machine Interface
• Shared Control
• Radio-Frecuency Identification
• Human-Robot Interface
• Robot Arm