The most important part of infrastructure is the road system since it directly affects people's lives by enabling connectivity and movement. Currently, the responsible authorities spend a significant amount of money using traditional methods to evaluate the state of the pavement. Smartphones, which have become a staple of modern life, can be used to evaluate the quality of the pavement. Researchers have already looked into the suitability of several smartphone sensors for assessing pavement condition. The precision and dependability of such methods haven't yet been studied, though. In the current exploratory work, an effort is made to investigate the applicability and accuracy of smartphone sensors in estimating pavement roughness. ProVAL double integration method, Quarter Car Simulation (QCS) method, magnitude of Fast Fourier Transform (FFT) method, and Power Spectral Density (PSD) analysis were four different smartphone sensor-based pavement roughness estimation techniques that were developed, and the technique that showed the best correlation with the roughness measurements obtained using a standard Roughometer was identified. When the host car is permitted to travel in three distinct speed ranges, namely 0–20 km/h, 21–40 km/h, and 41–60 km/h, the correlation between smartphone and Roughometer based roughness readings were examined for all four approaches. It was identified that the host vehicle speed range of 41–60 km/h resulted in the highest correlation value for all the four methods. The PSD analysis method was identified to be the most reliable one as it delivered the highest R2 value of 0.826.