Hi,
I took the time to actually build this to see if it would work, for a small motor about the size of a quarter in dia. runs quite fast at around 2.5 volts, across it.
This was designed to work as low as 3v. so 3.6v. will work fine.
I used a IR transistor, for the detector.
Here's how it works,
The motor has a DC resistance of around 12ohms.
The first stage (motor driver), is biased to have at the very least around 100mA. of collector current.
With this biasing and testing the motor runs very fast with 2.6v. dropped across it.
The second stage (detector), is biased to put its collector voltage at a value that would make the output transistor go into cutoff. At this point the motor shows 0v. dropped across it.
The I-R diode used in this experiment, was the kind you can get at radioshack, this was hooked up with a current limiting resistor of 470 ohms.
This diode was held at a distance of around 1" and turned on, the detector (I-R transistor) picked up this signal and brings the voltage at its collector low, thereby bringing the transistor invertor to a low base voltage cutting it off to the point where the output transistor motor driver can go into conduction to drive the motor to full voltage.
The voltage drop across the motor reached 2.6v. continuous running as long as the IR transmision was happening, this was very directional, with the tip of a pencil point I was able to break the IR beam very acurately and by moving the pencil point slowly different directions was able to vary the speed of the motor depending on the amount of IR light was recieved by the detector.
You can give it a try and see if you can get it working for yourself.
Hope this helps in some way...