The Brain Machine Interface (BMI) uses a combination of electrical and heat scanners to identify activity in the brain, and sophisticated pattern-identification software to match that activity with specific patterns of thought it can translate into simple commands.

As a demonstrator for the interface, one version of Honda’s humanoid Asimo robot was configured to respond to control signals it receives from the BMI setup using a wireless data connection.

This BMI is a second for Honda which developed one in 2006 that was based on Functional Magnetic Resistance Imaging (fMRI), which creates images of the brain by monitoring changes in a magnetic field as it passes through the head. fMRI is designed to create images in real time to help identify changes specific to a pattern of thought.

The most recent version – which the Honda Research Institute built with help from the Advanced Telecommunications Research Institute International and Shimadzu Corp. – uses a combination of near-infrared spectroscopy (NIRS) combined with more traditional electroencephalography (EEG).

NIRS is a form of optical topography, which is widely used as a way to map brain functions. There has been substantial research into the use of BMI in medical and surgical applications, though this proof-of-concept is able to allow a child to control a model train by brainwaves.

Most are designed to help disabled or injured patients communicate using computers, though the Defense Advanced Research Projects Agency has sponsored a number of BMI research projects designed to assist healthy people as well.

Experiments on monkeys have demonstrated it’s possible to create a higher-functioning connection with implanted electronics that allow the brain to actually treat a robotically assisted appendage to function like a real one.

Honda did not announce for what other applications it might apply the BMI technology.