Learning from Demonstration for Autonomous Navigation in Complex Unstructured Terrain

By David Silver, J. Bagnell, Anthony Stentz. 2010

183 citations recorded by the source corpus.

Rough terrain autonomous navigation continues to pose a challenge to the robotics community. Robust navigation by a mobile robot depends not only on the individual performance of perception and planning systems, but on how well these systems are coupled. When traversing complex unstructured terrain, this coupling (in the form of a cost function) has a large impact on robot behavior and performance, necessitating a robust design. This paper explores the application of Learning from Demonstration to this task for the Crusher autonomous navigation platform. Using expert examples of desired navigation behavior, mappings from both online and offline perceptual data to planning costs are learned. Challenges in adapting existing techniques to complex online planning systems and imperfect demonstration are addressed, along with additional practical considerations. The benefits to autonomous performance of this approach are examined, as well as the decrease in necessary designer effort. Experimental results are presented from autonomous traverses through complex natural environments.

Subfields: Aerospace Engineering, Computer Vision and Pattern Recognition, Control and Systems Engineering

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Aerospace Engineering

In Computer Vision and Pattern Recognition

In Control and Systems Engineering

Citations in the corpus

Reinforcement learning in robotics: A survey (2013, methodology): “These techniques, and many variants, have been recently successfully applied to outdoor robot navigation (Ratliff et al., 2006a; Silver et al., 2008, 2010), manipulation (Ratliff et al., 2007), and quadruped locomotion (Ratliff et al., 2006a, 2007; Kolter et al., 2007).”
Driving Policy Transfer via Modularity and Abstraction (2018, background): “ive to hand-designed modular pipelines. Robotic vehicles equipped with neural network policies trained via imitation learning have been demonstrated to perform lane following [37,3], off-road driving [25,44], and navigation in simple urban environments [8]. However, training these methods in the phy”
End-to-End Driving Via Conditional Imitation Learning (2017, background): “Imitation learning has been applied to a variety of tasks, including articulated motion [2], [28], [11], autonomous flight [1], [13], [30], modeling navigational behavior [37], [38], off-road driving [22], [32], and road following [4], [6], [27], [36].”
CARLA: An Open Urban Driving Simulator (2017, background): “The development of autonomous ground vehicles is a long-studied instantiation of this problem [22, 26].”
An Algorithmic Perspective on Imitation Learning (2018, methodology): “Ratliff et al. [2006b], Silver et al. [2010] apply maximum margin planning (MMP) and LEARCH for finding a path with minimum accumulated cost (see Figure 4.3).”
CIRL: Controllable Imitative Reinforcement Learning for Vision-based Self-driving (2018, background): “As a straight-forward idea, imitation learning that learns policies via supervised training on human driving data has been applied to a variety of tasks, including modeling navigational behavior [20], off-road driving [12,2], and road following [9,8,1,18,11].”
Planning and Decision-Making for Autonomous Vehicles (2018, background): “A similar variant, maximum margin planning (117), was applied to navigate a robot in complex unstructured terrain (118) and to learn autonomous driving styles and maneuvers (119).”
A survey of inverse reinforcement learning techniques (2012, methodology): “Then the LEARCH algorithm was implemented in an autonomous navigation problem where the vehicle was operated in a complex unstructured terrain (Silver et al., 2010).”
Learning by Cheating (2019, methodology): “It was pioneered by Pomerleau [17] and developed further in subsequent work [5, 12, 21].”
Learning Ground Traversability From Simulations (2017, background): “For example, in [24] a human draws a path from a source to a target, and the system infers which areas the human purposefully avoided and automatically uses such patches as”
Recent Advances in Robot Learning from Demonstration (2020, methodology): “To arrive at a unique reward function, IRL methods consider different additional optimization goals, such asmaximummargin (112, 117, 121, 122) and maximum entropy (58, 118, 123, 124).”
Active learning from demonstration for robust autonomous navigation (2012, methodology): “In order to evaluate the effectiveness of both novelty and uncertainty based active learning, experiments were performed in the context of learning from demonstration for interpreting overhead data of an environment [1], [20].”