This work presents a simple model that expose the information embedded into a sensor signal allowing to share it independently of the signal nature. Today's highly interconnected world requires a representation of sensor signals that let efficient sharing of embedded information. The proposed model is a state matrix that combine two important aspects of any signal: Its value inside a range and its behavior over the time. From this state matrix is possible to obtain a self-learning model observing the state transition probabilities and the time lapse in each state to deduce signal normality- abnormality that allows to infer a better perception of reality.
