Machine learning applications in off-road vehicles interaction with terrain: An overview

Abstract

With the advent of artificial intelligence, the analysis of systems related to complex processes has become possible or easier. The interaction of the traction factor of off-road vehicles with soil or other uncommon surfaces is one of the complex mechanical problems, which has been very difficult to model and analyze in conventional and previous methods due to numerous and variable parameters. This review article delves into the imperative and progression of integrating AI algorithms within the realm of modeling and predicting target parameters in Terramechanics engineering. Such endeavors are especially pertinent to predicting various soil properties, including soil compaction, traction, energy consumption, deformation, and associated factors. The application of AI encompasses various facets, including modeling and predicting traction, soil sinkage, rut depth, contact area, soil stress, density, and energy wasted on the traction device’s movement on the soil. The present study evaluates the solutions and benefits offered by AI-based methodologies in addressing soil-machine interaction challenges. Furthermore, the study investigates the constraints inherent in utilizing these methodologies.