Characterizing Soil Stiffness Using Thermal Remote Sensing and Machine Learning

Soil strength characterization is essential for any problem that deals with geomechanics, including terramechanics/terrain mobility. Presently, the primary method of collecting soil parameters through in situ measurements but sending a team people out to site collect data this has significant cost implications and accessing location necessary equipment can be difficult. Remote sensing provides an alternate approach measurements. In lab study, we compare use Apparent Thermal Inertia (ATI) against GeoGauge direct testing stiffness. ATI correlates stiffness, so it allows one predict remotely using machine-learning algorithms. The best performing regression algorithm among ones tested different predictor variable combinations was found KNN R2 0.824 RMSE 0.141. This study demonstrates potential remote acquire thermal images characterize terrain mobility utilizing