Electronic Thesis and Dissertation Repository

Thesis Format



Master of Engineering Science


Electrical and Computer Engineering

Collaborative Specialization

Planetary Science and Exploration


McIsaac Kenneth


An instrumented rover wheel can collect vast amounts of data about a planetary surface. Planetary surfaces are changed by complex geological processes which can be better understood with an abundance of surface data and the use of terramechanics. Identifying terrain parameters such as cohesion and angle of friction hold importance for both the rover driver and the planetary scientist. Knowledge of terrain characteristics can warn of unsafe terrain and flag potential interesting scientific sites. The instrumented wheel in this research utilizes a pressure pad to sense load and sinkage, a string potentiometer to measure slip, and records motor current draw. This thesis demonstrates the utilization of the instrumented wheel's data to estimate cohesion, angle of friction and grain size and demonstrates a machine learning solution for classifying terrain types with the same data. Mars simulants available at NASA-JPL were used for the collection of the data. Two machine learning classifiers were explored: Random Forest and Support Vector Machine. Binary and multi-class classification were both demonstrated and it is proposed that the classification model can identify terrain types based on the instrumented wheel data. The Random Forest model performed best in all classification types.

Summary for Lay Audience

Terramechanics is the study of vehicle-terrain interaction. Terrain characteristics change dependent on geologic processes such as erosion that may have occurred in a region. A change in terrain characteristics can be observed via terramechanics models which define the interaction of vehicles in the terrain. Through the collection of data on the interaction of a planetary rover with the surface terrain, terramechanics models can be used to estimate terramechanics parameters which define terrain characteristics. The identification of these parameters is important for both the rover driver and planetary scientist as they reveal if the terrain is unsafe for driving, previously unencountered and/or of scientific interest, and reveal the history of the planetary region where the rover finds itself. Cohesion, Angle of friction, and grain size are important parameters that can give insight into the safety for driving or history of the region. This thesis demonstrates the capability of a "smart" sensing rover wheel to collect surface terrain data and the use of this data in estimating terrain parameters, and the identification of terrain types using machine learning methods. Two classification algorithms were explored: Random Forest and Support Vector Machine. It is proposed that both binary and multi-class classification models can differentiate between Mars simulants used to collect data for this study using pressure pad, wheel slip, and motor current data.