Current Transformers are really only usable for AC currents.
Placing a resistor in line and measuring voltage drop is great in theory, but the currents we deal with are rather high. 120A through a milliohm is 14 watts, and your robot will draw several times that routinely (for relatively short time scales). Some teams have taken to using their existing wiring as a series resistance and met with success. As the wire heats, you will lose about half a percent of accuracy per Kelvin (ish). Likely doesn't matter for this application.
I suggest using a hall effect sensor. Allegro Microsystems make some really nice in-line ones:
http://www.allegromicro.com/en/Produ...ensor-ICs.aspx
Melexis makes one that I've used before with mixed results:
http://www.melexis.com/Hall-Effect-S...SA-1V-614.aspx
For this, you "simply" place the wire you want to measure near the sensor and measure the magnetic field. Your accuracy will then depend largely on placement. Great for relative readings though! It also doesn't wrap a flux guide around the current source, so it may be more prone to external noise sources than packaged solutions.
There are also things that look like current transformers, but actually have a hall effect sensor and a magnetic flux guide inside them. I don't have a part number on hand for those right now. Very nice, but typically pricier.
Be careful to read the electrical rules as to which might be legal.
Also, you might want to consider the benefit gained by the system. If you haven't popped the breaker yet, you may just be hobbling your capabilities.