• Corpus ID: 14814869

Self-Supervised Terrain Classification for Planetary Rovers

@inproceedings{BrooksSelfSupervisedTC,
  title={Self-Supervised Terrain Classification for Planetary Rovers},
  author={Christopher A. Brooks and Karl Iagnemma}
}
—For planetary rovers, autonomous mobility is a key to enabling greater science return. While current terrain sensing approaches can be used to detect geometric hazards, such as rocks and cliffs, they are limited in their ability to detect non-geometric hazards, such as loose sand in which a rover may become entrenched. This paper presents a self-supervised classification approach to learning the visual appearance of terrain classes which relies on vibration-based sensing of wheel-terrain… 

Figures and Tables from this paper

Self-supervised terrain classification based on moving objects using monocular camera
TLDR
This paper presents a self-supervised classification approach to enable a robot to learn the visual appearance of terrain classes in various outdoor environments by observing moving objects and to learn about the terrain, based on their paths of movement.
Adaptive mobile robot navigation based on online terrain learning
TLDR
This work aims at establishing a framework from which the rover can learn from its interaction with the terrain it encounters and shows the importance of such a method.
Novelty Detection and Online Learning for Vibration-based Terrain Classification
TLDR
This paper presents a vibration-based terrain classification system that uses novelty detection based on Gaussian mixture models to detect if the robot traverses an unknown terrain class and adds a new terrain class to the classification model online.
Adaptive rover behavior based on online empirical evaluation: Rover–terrain interaction and near‐to‐far learning
TLDR
This work aims at establishing a framework from which the rover can learn from its interaction with the terrains encountered and shows the importance of such a method, which is called rover–terrain interactions learned from experiments (RTILE).
Adaptive rover behavior based on online empirical evaluation: Rover–terrain interaction and near-to-far learning
TLDR
This work aims at establishing a framework from which the rover can learn from its interaction with the terrains encountered and shows the importance of such a method, which is called rover–terrain interactions learned from experiments (RTILE).
A combination of vision- and vibration-based terrain classification
TLDR
This paper presents a terrain classification approach which fuses terrain predictions based on image data with predictions made by a vibration-based method, and shows that by fusing both predictions, the classification rates are significantly larger than predicted based on data from a single sensor alone.
In-situ soil sensing for planetary micro-rovers with hybrid wheel-leg systems
Rover missions exploring other planets are tightly constrained regarding the trade-off between safety and traversal speed. Detecting and avoiding hazards during navigation is capital to preserve the
Development of an autonomous robot for ground penetrating radar surveys of polar ice
TLDR
The design and fabrication of a low cost, battery-powered mobile robot for ground penetrating radar surveys in support of Polar science and logistics and a proprioceptive sensor suite and sample data for autonomous detection of terrain traversability are described.
Self-Supervised Sensor Learning and Its Application: Building 3D Semantic Maps Using Terrain Classification
TLDR
A method that enables a robot to classify a terrain in various outdoor environments using terrain information that it recognizes without the assistance of a user is proposed; then, it creates a three-dimensional (3D) semantic map.
Self-learning classification of radar features for scene understanding
...
1
2
...

References

SHOWING 1-10 OF 27 REFERENCES
Self-Supervised Classification for Planetary Rover Terrain Sensing
TLDR
A self-supervised classification approach to learning the visual appearance of terrain classes which relies on vibration-based sensing of wheel-terrain interaction to identify these terrain classes is proposed.
On-line terrain parameter estimation for planetary rovers
TLDR
Simulation and experimental results show that the terrain estimation algorithm can accurately and efficiently identify key terrain parameters for loose sand, which are valuable indicators of planetary surface soil composition.
Terrain estimation for high-speed rough-terrain autonomous vehicle navigation
High-speed unmanned ground vehicles have important potential applications, including reconnaissance, material transport, and planetary exploration. During high-speed operation, it is important for a
Traversability classification using unsupervised on-line visual learning for outdoor robot navigation
TLDR
A novel on-line learning method which can make accurate predictions of the traversability properties of complex terrain based on autonomous training data collection which exploits the robot's experience in navigating its environment to train classifiers without human intervention.
Visual wheel sinkage measurement for planetary rover mobility characterization
TLDR
A vision-based method to measure the sinkage of a rigid robot wheel in rigid or deformable terrain using a single grayscale camera and is computationally efficient, making it suitable for systems with limited computational resources such as planetary rovers.
TERRAIN ESTIMATION USING INTERNAL SENSORS
TLDR
A novel algorithm is presented that uses the vehicle’s internal sensors to qualitatively categorize the terrain type in real-time and was successful in identifying gravel, packed dirt, and grass.
Stereo vision and rover navigation software for planetary exploration
TLDR
The radiation effects analysis is summarized that suggests that commercial grade processors are likely to be adequate for Mars surface missions, and the level of speedup that may accrue from using these instead of radiation hardened parts is discussed.
Terrain parameter estimation and traversability assessment for mobile robots
The estimation of terrain characteristics is an important missions of Martian exploration rovers. Since only limited resources and human supervision are available, efficient and autonomous method of
Wheel slippage and sinkage detection for planetary rovers
Mobile robots are increasingly being used in high-risk rough terrain situations, such as planetary exploration and military applications. Current control and localization algorithms are not well
Learning slip behavior using automatic mechanical supervision
TLDR
A novel probabilistic framework in which the visual information and the mechanical supervision interact to learn particular terrain types and their properties is presented, which will enable the rover to drive safely on slopes, learning autonomously about different terrains and their slip characteristics.
...
1
2
3
...