The abilities of the window module are very basic - just wind the windows up and down, when the ignition is on. Not much else. But WE WANT MORE :)
So it has been suggested that maybe a system could be implemented to wind the windows up when the doors are locked via the remote.
The window control module already has a timer circuit built in which allows the front windows to wind up with just a tap on the button. We can use this function to be triggered by a pulse from a micro to wind the window up.
The next step will be to find the right signal to trip the window UP command.
I suspect there will be a lock command on the I-Bus which could be used or it might be possible to take a signal direct from the from the TWICE module as it is located under one of the front seats.
Having a poke around in the TWICE module I would be looking for the output from the remote receiver and seeing if I can easily differentiate between the lock and unlock signal. This must be a signal that is only present when activated via the keyfob and not say a lock button in the car or the auto lock function.
If Cracking the TWICE module open proves to distressing the I Bus may have what We need as well but with added complexity of decoding and filtering CAN data.
Some usefull information on the I Bus I have found at http://pikkupossu.1g.fi/tomi/projects/i-bus/i-bus.html#720 while specifically for a 9-3 I would take a punt that there would be sufficient similarities for the data to be useful.
We would need to provide power as well to the window module as its power is cut when the ignition is turned off. Power for the Central locking is routed through the TWICE module which is supplied by a 30A fuse which would be possibly be suitable for the windows - maybe with a slight delay to activate the window after the doors have locked to prevent excess current draw.