Tires with embedded sensors enable them to be active sensing components. Intelligent tires have increased potential in sensing not only the state of the tire itself but the related properties of the road. This Technical Note briefly describes the application of accelerometer-based intelligent tires in sensing the adhesion level and unevenness of the road. A combined tire longitudinal dynamic model is introduced to estimate the tire-road friction from the waveform features of the acceleration signals. Data-driven methods are adopted to directly classify the road unevenness level from the original signals. Primary analyses and experiments verify the proposed method. Road sensing with intelligent tires contributes to the development of active safety control and driving assistance systems.