Soil surface roughness (SSR) is strongly influenced by agricultural activities and soil properties. However, measuring SSR is tedious and time-consuming, especially under stationary conditions. This work was conducted with laboratory measurements of SSR to investigate how to measure it on-the-go. Four roughness indices were examined to investigate which one provides a more representative value for measuring SSR on-the-go. Out of all indices, random roughness (RR) was selected for the next stage of the study, since it refers to the degree of fineness and pulverization of the soil. Four roughness classes of A, B, C, and D were formed with depressions of 0.5, 1, 2, and 4 cm, respectively, on the soil surface using four corrugated steel sheets. The RR was calculated for these corrugated soil surfaces using two range-finder sensors, an infrared (IR) and an ultrasonic (Ult) type. A carrier was built up to move the range-finder sensors above the corrugated soil surfaces at travel speeds of 2.6, 3.5, and 4.8 km/h. The overall averages of RR determined for IR and Ult sensors were correlated with the reference values measured using a pin-type profiler with correlation coefficients of 0.99 and 0.31, respectively. No significant differences were observed between the calculated RRs at the examined travel speeds. The depression and peaks measured using IR and Ult range-finder sensors were also compared to the reference values measured using the pin-type profiler and the results showed average correlation coefficients of 0.97 and 0.69 for IR and Ult range-finder sensors, respectively. While both range-finder sensors could thoroughly measure the distance under stationary conditions (R2>0.99), however, only the IR sensor succeeded in measuring RR on-the-go on roughness classes A, B, C, and D. The reason may be attributed to the wider footprint of the Ult sensor. The results of this study can help predict physical properties related to SSR more quickly.
