choosing motor type

I'm really green when it comes to power circuits and motors so please bare with me cause I need a little direction.

I want to use an electric motor to shift a motorcycle. A motorcycle has a sequential transmission and a rotating shaft that shifts up when you rotate it ~30 degrees CCW from the neutral position and downshifts when you rotate CW ~30 degrees from the neutral position. It has a 1 inch lever arm on it but that can be removed. It requires about 50lb-in of torque to rotate it, but I want to be able go from neutral to fully engaged in less than 100ms. My control computer has a 15A output limit on it's high power pins, but it wouldn't be difficult for me to rig up a mosfet that I control with the ECU and can pass higher current if the motor needs it.

Other important information is the battery can supply 6.9A/hour, 350CCA, andis 12VDC. The package needs to be relatively small, so about the size of a fist, maybe a little larger.

so now to the questions
0: is the sub 100ms performance possible given the constraints I've listed?

1: What type of motor would you use for this?
I was thinking a stepper motor because of it's ability to hold discrete positions and the high initial toruqe, but that's just from my initial (and limited research)

2:What's the method to control this? My ECU can run a simulink model for the control scheme (converted to C), and it can vary voltage but not current from it's 15A output pins. I'm just looking for a general understanding of how it would work and I can work out the nitty gritty later.

Thanks for the help.

You say it upshifts when rotated CCW _from_ neutral and downshifts when rotated CW _from_ neutral. So what if you want to continue to upshift? Is it going to be like CCW 30 then back to zero (neutral) and then again CCW 30 OR CCW 30 and hold for continuous upshifting and then to neutral to hold the gear? Similarly for downshifting.

Whatever be the case, it looks like you need a bidirectional motor - not sure if a stepper motor is bidirectional also. Now I am no expert at this, but instead of a motor you could use a linear mechanism to rotate the shaft (connect it via a joining rod to the shaft) and have the plunger move back and forth - a couple of solenoids can be used to position the plunger and the return to neutral can be via a spring.

Sorry I wasn't more clear in the OP, after every shift, the arm must return to the neutral position before you can shift again, either up or down. You can't "overshift" into another gear, so you can only go up or down one gear at a time, so if you applied something absurd like 20Nm of torque, you'll just bottom out the rotary shaft until you relieve the force, at which point it will spring back to neutral (the transmission has rotary shaft has an internal spring).

But yes, it must be bi-directional. I don't think a linear actuator would work, I considered that before but it was just too slow for the force I needed. Currently I used an electronically controlled pneumatic system for just that reason.

John,
Can you relax your time spec? It would make the solution much simpler and less expensive.