de problem is hen de day is over de solar panel mast go beck to bi preperd to de next day.
sorry my ingles is verry bad 
I can use light sensors to move the engine or any timer.
that you couple the idea.
the problem is that at the end of the day, the engine has to return to your starting point.
sei and the arduino can help me 
It's not clear what you will be moving, or how you will move whatever it is you will be moving.
But, if you are moving a solar panel, using servos, with light sensors to follow the sun, the Arduino can handle that.
Track the sun all day. When the light diminishes at the end of the day, below some threshold for a sufficient period of time, the servos should return the panel to the start position, ready for the next day.
You'll need to be careful to wait a while after the light level drops, to account for thunderstorms during the day.
thank you very much ;),
that's the idea and ;D happy to be able to use arduino.
please be some I can make a list of what I have to buy would be very grateful
I hope soon to put the results here.
thank you very much
Well, you'll need an Arduino.
You'll need some means of detecting light levels. There are a number of different kinds of light detecting sensors.
You'll need something to move. You haven't said what that is.
You'll need some way to hold the thing to be moved.
You'll need some motivating device(s) to move whatever needs to be moved. This can be servos, stepper motors, solenoids, gear motors, pneumatic controls, hydraulic controls, etc. Depends on what fabrication skills you have, how you want to move whatever it is needs to be moved, and how heavy that thing is.
Some means of powering the Arduino and sensors would be useful.
A few wires to connect stuff together.
Some lights to indicate whether stuff is working might be good. If that includes LEDs, some resistors will be useful.
Diodes, transistors, etc. may be needed, depending on what you need to control. Maybe an H-bridge or motor controller.
Really, you've told us too little about your project or where you are or your skill levels for us to be putting together a shopping list for you.
In the 70's, I made a 'sun tracking' solar collector using a very slow motor (mains induction motor & gearbox came out of a spit roast, and then geared down to 1 rev in 24 hours-ish).
Wasn't very accurate, and ran all day (used about 4W/hour).
Captured about 300-500W on a sunny (British) day. No microcontroller.
I would think that another method might also be interesting in driving a solar seeker panel. Rather then having sensor feedback, how about having a on board real time clock module and a large look up table to determine when to move and when to reset at the end of the day. The table data would have to be calculated for the latitude and date of the local position, but would not have to worry about changing light brightness, passing storms, etc. With the large flash program space in the mega one could probably fit all the table look-up data using: Flash | Arduiniana
Seems to me that could say a lot of cash and mechanical design problems and work. A simple stepper motor could drive a lead screw to drive the panel through it's total range, with just possibly needing limit switches at the ends of travel as references and safety stops.
Any thoughs?
Lefty
Retrolefty - you definitely 'got' the story.
Assuming the mechanical setup is appropriate, I would add ...
A simple time clock-based approach, moving a solar collector only once per hour, is never going to be more than +/-7.5 degrees misaligned East/West.
If it were moved every 12 minutes (1/5th of an hour), the error would be less than +/-1.5 degrees.
The loss of efficiency due to a solar panel being misaligned by 1.5degrees is likely to be small, and maybe less than the variability of solar panels (I'd drive my calculation about how often to move the panel from that sort of number)
North/South misalignment only needs to be fixed once/day, at most, using a calendar (assuming the axis of East/West rotation is the thing being adjusted, like a telescope equatorial mount).
It would probably be quite difficult to make a sun-light-sensitive tracking system which yields noticeably better results.
I agree steppers, and end stop sensors would be good. Highly geared little DC motors with encoders would probably be fine with this sort of duty cycle.
As PaulS said, we need to understand a bit more about the size and nature of the solar panel.
HTH
GB
PS - According to Diurnal motion - Wikipedia a Sun diameter is moved every 2 minutes.
thank you very much everybody for help me,
panels are large is an autonomous system is 12v.
I have not experienced programming with which my level is zero, jejeje
with great gusto.
within days hardmeeting in Barcelona and I can decover this world looking for your answers very thanks
SOStainable health