jamauro18:
Right now we are using time because the actuator does not send location information.
That isn't ideal - without any feedback you won't get accurate positioning, and if you're allowing movement between positions these errors would be cumulative. You might actually be better off bringing the actuator back to a home position before each movement, as you've already described.
If you definitely want to allow relative movements then you need to remember the current position of the actuator and subtract that from the desired end position to work out how far to move it and in which direction. Then you can use your timed method to move the actuator.