> we would suggest illuminance tests followed by normalizations of the data to observe deviations under various situations.

I haven't seen anything on this while working in embedded space. If anyone can throw more light on this. Like what tests should go where, and how to make sure that they are successful.