The dynamic parameters of a roller rig vary as the adhesion level changes. The change in dynamics parameters needs to be analysed to estimate the adhesion level. One of these parameters is noise emanating from wheel–rail interaction. Most previous wheel–rail noise analysis has been conducted to mitigate those noises. However, in this paper, the noise is analysed to estimate the adhesion condition at the wheel–rail contact interface in combination with the other methodologies applied for this purpose. The adhesion level changes with changes in operational and environmental factors. To accurately estimate the adhesion level, the influence of those factors is included in this study. The testing and verification of the methodology required an accurate test prototype of the roller rig. In general, such testing and verification involve complex experimental works required by the intricate nature of the adhesion process and the integration of the different subsystems (i.e. controller, traction, braking). To this end, a new reduced-scale roller rig is developed to study the adhesion between wheel and rail roller contact. The various stages involved in the development of such a complex mechatronics system are described in this paper. Furthermore, the proposed brake control system was validated using the test rig under various adhesion conditions. The results indicate that the proposed brake controller has achieved a shorter stopping distance as compared to the conventional brake controller, and the brake control algorithm was able to maintain the operational condition even at the abrupt changes in adhesion condition.