Low adhesion between railway wheel and rail, which is usually induced by contaminants such as water, oil, leaves etc., affects railway operation in terms of performance and safety. This study uses an optical sensor to identify different surface layers which cause low adhesion. A laboratory set up and field tests under various conditions were subject to the surface layer identification by the optical sensor. In addition, the friction coefficient was measured on the same surface layers. The results show that the sensor can distinguish between different surface layers. This information further linked to the levels of the friction coefficient, which can be used in the prediction system for the railway operator.