In this paper an Extended Kalman Filter (EKF) is used as an observer for temperature monitoring, like a virtual sensor, of a metal-polymer fibre based heater structure. Metal-coated polymers are relevant for the realisation of smart systems (capable of both sensing and actuating). A real-time implementation of the temperature estimator is important to guarantee a gentle, fault-free operation of the heater and to reject disturbances. It can be used to control the resulting temperature without a direct measurement or to guarantee the users' safety by reacting to overheating. An estimation strategy is necessary because for normal operation in the real world, measurements of the temperature are usually not available, but it is important to know it in order to implement health monitoring algorithms or to do a safe shutdown if necessary. Simulations justify the observation strategy while measurements validate the utilized model.