We have developed a brain-machine interface to control the speed of an RC car. We are using electroencephalogram (EEG) signal measured from the the brains of human operators to close the loop. The EEG signals are measured using Mindflex headsets. We have developed a control program on an Atmel microcontroller to achieve control.


  • We have cracked open a receiver base and located a PWM signal used for the Mind Flex fan motor. We use an edge detection algorithm in our Atmel microcontroller to measure the duty cycle of this signal. Unfortunately only one Mindflex device can be used at a time. We used this signal to produce speed control. We also tried to use this signal for turning with a constant speed, but it was too difficult to a control the signal finely enough to produce turning.


  • We used Atmel attiny261 microcontrollers on a hand-build project board with an in-circuit programmer to adapt observable brain signals into machine control signals. The fan control signal from the mindflex devise is a pulse width modulated signal with six steps between 0% and 95% duty cycle. This duty cycle is translated into a control signal sent to the Duratrax Sprint RC motor controller.

I this image P varies between 1ms and 2ms where 1.5ms is idle. The Duratrax motor controller requires W=20ms period of the pulse width modulation signal.


This result video shows the speed of the car being controlled via the mindflex headset. Turning is done by a second operator using a Duratrax transmitter. [bmi_car.avi Download]