Vibration-based terrain classification for planetary exploration rovers

Brooks, Christopher; Iagnemma, Karl

doi:10.1109/tro.2005.855994

Cited by 196 publications

(146 citation statements)

References 13 publications

Supporting

Mentioning

144

Contrasting

Order By: Relevance

“…features derived from sensor data measuring some aspect of physical robot-terrain interaction. A method for terrain classification based on analysis of vibrations arising from robot wheel-terrain interaction was first proposed in (Iagnemma and Dubowsky, 2002) and developed by Brooks and Iagnemma (2005).…”

Section: Related Workmentioning

confidence: 99%

“…Analysis of vibrations propagating through a rover's wheel/suspension structure can be used to distinguish between various types of terrain the rover is traversing (Brooks and Iagnemma, 2005). This classification mode is unique among the low-level classifiers described here in that it relies on a "tactile" sensor signal that is modulated by physical rover-terrain interaction.…”

Section: Vibration Feature Selectionmentioning

confidence: 99%

“…The general classification framework employed here is identical to that in (Brooks and Iagnemma, 2005). Vibration signals were processed as the log power spectral density for every one-second time step at 557 frequencies in the frequency range 20.5 Hz to 12 kHz.…”

Section: Vibration Feature Selectionmentioning

confidence: 99%

“…A linear classifier was trained in this 557-dimensional space using hand-labeled training data. For this work, a support vector machine with a linear kernel was used as the classifier, instead of the Fisher linear discriminant employed in (Brooks and Iagnemma, 2005).…”

Section: Vibration Feature Selectionmentioning

confidence: 99%

See 3 more Smart Citations

A study of visual and tactile terrain classification and classifier fusion for planetary exploration rovers

2008

View full text Add to dashboard Cite

Knowledge of the physical properties of terrain surrounding a planetary exploration rover can be used to allow a rover system to fully exploit its mobility capabilities. Terrain classification methods provide semantic descriptions of the physical nature of a given terrain region. These descriptions can be associated with nominal numerical physical parameters, and/or nominal traversability estimates, to improve mobility prediction accuracy. Here we study the performance of multi-sensor classification methods in the context of Mars surface exploration. The performance of two classification algorithms for color, texture, and range features are presented based on maximum likelihood estimation and support vector machines. In addition, a classification method based on vibration features derived from rover wheel-terrain interaction is briefly described. Two techniques for merging the results of these "low level" classifiers are presented that rely on Bayesian fusion and meta-classifier fusion. The performance of these algorithms is studied using images from NASA's Mars Exploration Rover mission and through experiments on a four-wheeled test-bed rover operating in Mars-analog terrain. Also a novel approach to terrain sensing based on fused 2 tactile and visual features is presented. It is shown that accurate terrain classification can be achieved via classifier fusion from visual and tactile features.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Vibration Feature Selectionmentioning

confidence: 99%

Section: Vibration Feature Selectionmentioning

confidence: 99%

Section: Vibration Feature Selectionmentioning

confidence: 99%

See 2 more Smart Citations

A study of visual and tactile terrain classification and classifier fusion for planetary exploration rovers

2008

View full text Add to dashboard Cite

show abstract

“…Brooks and Iagnemma presented vibration-based terrain classification for low-speed planetary rovers [3]. They use Principal Component Analysis (PCA) to reduce the dimensionality of the vibration data and Linear Discriminant Analysis (LDA) for classification.…”

Section: Introductionmentioning

confidence: 99%

SVMs for Vibration-Based Terrain Classification

Weiß

Stark

Zell

2007

Autonome Mobile Systeme 2007

View full text Add to dashboard Cite

Abstract. When an outdoor mobile robot traverses different types of ground surfaces, different types of vibrations are induced in the body of the robot. These vibrations can be used to learn a discrimination between different surfaces and to classify the current terrain. Recently, we presented a method that uses Support Vector Machines for classification, and we showed results on data collected with a hand-pulled cart. In this paper, we show that our approach also works well on an outdoor robot. Furthermore, we more closely investigate in which direction the vibration should be measured. Finally, we present a simple but effective method to improve the classification by combining measurements taken in multiple directions.

show abstract

State and Unknown Terrain Estimation for Planetary Rovers via Interval Observers

Khajenejad

Jin

Yong

2021

Advanced Intelligent Systems

View full text Add to dashboard Cite

Herein, the problem of state and unknown terrain estimation is considered, where the unknown planetary terrain parameters, e.g., terrain stiffness and ground height, are inferred from how it affects rover motion through vehicle‐terrain interaction. In particular, an alternative framework for terrain estimation based on set‐valued or set‐membership estimation is proposed, where the goal is to find set‐valued estimates (in the form of hyper‐rectangles or intervals) of the states and unknown terrain parameters. For this purpose, a state and model interval observer is designed for partially unknown nonlinear systems with bounded noise. By leveraging a combination of nonlinear bounding/decomposition functions, affine abstractions, and a data‐driven function abstraction method (to overestimate the unknown dynamics model from noisy input–output data), the proposed observer is capable of simultaneously estimating the states and learning the unknown dynamics. Further, a tractable sufficient condition is derived for guaranteeing the stability of the designed observer, i.e., such that the sequence of interval estimate widths are uniformly bounded. When applied to the state and unknown terrain estimation problem, the simulation results indicate that our approach can more reliably find the range of possible terrain parameters when compared with the cubature Kalman filter.

show abstract

Vibration-based terrain classification for planetary exploration rovers

Cited by 196 publications

References 13 publications

A study of visual and tactile terrain classification and classifier fusion for planetary exploration rovers

A study of visual and tactile terrain classification and classifier fusion for planetary exploration rovers

SVMs for Vibration-Based Terrain Classification

State and Unknown Terrain Estimation for Planetary Rovers via Interval Observers

Contact Info

Product

Resources

About