From perception to decision: A data-driven approach to end-to-end motion planning for autonomous ground robots

By Mark Pfeiffer, Michaela Schaeuble, Juan I. Nieto, Roland Yves Siegwart, César Cadena. 2016

428 citations recorded by the source corpus.

Learning from demonstration for motion planning is an ongoing research topic. In this paper we present a model that is able to learn the complex mapping from raw 2D-laser range findings and a target position to the required steering commands for the robot. To our best knowledge, this work presents the first approach that learns a target-oriented end-to-end navigation model for a robotic platform. The supervised model training is based on expert demonstrations generated in simulation with an existing motion planner. We demonstrate that the learned navigation model is directly transferable to previously unseen virtual and, more interestingly, real-world environments. It can safely navigate the robot through obstacle-cluttered environments to reach the provided targets. We present an extensive qualitative and quantitative evaluation of the neural network-based motion planner, and compare it to a grid-based global approach, both in simulation and in real-world experiments.

Subfields: Aerospace Engineering, Computer Vision and Pattern Recognition

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

Citations in the corpus

A Survey of Deep Network Solutions for Learning Control in Robotics: From Reinforcement to Imitation (2016, positive): “In robotics, the formulation of IRL offers an efﬁcient solution for learning policies for socially compliant navigation (Okal and Arras, 2016; Pfeiffer et al., 2016; Kretzschmar et al., 2016), where the agent needs to not only avoid collisions with static obstacles but also to behave in a socially…”
Spatial Action Maps for Mobile Manipulation (2020, methodology): “t all navigation systems based on ConvNets consider only a small set of (possibly parameterized) egocentric steering commands (e.g., move forward/backward, strafe left/right, rotate left/right, etc.) [26,20,11,12,24,47,29,7]. Representing actions with steering commands (Fig.2a) presents several prob”
ANYmal parkour: Learning agile navigation for quadrupedal robots (2023, background): “The locomotion module contains a catalog of locomotion skills that can overcome specific terrains.”
Lifelong Navigation (2020, background): “As for physical robot navigation, learning approaches typically either imitate a conventional planner [19] or learn from trial-and-error using reinforcement learning [3].”
DroNet: Learning to Fly by Driving (2018, background): “network with a rough estimate of the distance to the goal [26], or, if a coarse 2D map of the environment is available, exploit recent learning-based approaches developed for ground robots [27].”
Learning Generalized Reactive Policies using Deep Neural Networks (2017, positive): “Imitation learning has been previously used with DNNs to learn policies for tasks that involve short horizon reasoning such as path following and obstacle avoidance (Pomerleau 1989; Ross, Gordon, and Bagnell 2011; Tamar et al. 2016; Pfeiffer et al. 2016), focused robot skills (Mülling et al.”
Embodied Question Answering (2017, background): “Visual navigation typically speciﬁes agent goals either implicitly via the reward function [18, 19] (thus training a separate policy for each goal/reward), or explicitly by conditioning on goal state representations [23] like images of target objects [17, 24].”
Deep-learning in Mobile Robotics - from Perception to Control Systems: A Survey on Why and Why not (2016, methodology): “Pfeiffer et al. (2016) directly added target coordinates to the fully connected layer paralleling to the sensor feature extraction through CNN .”
A Survey of State-Action Representations for Autonomous Driving (2018, methodology): “The representation choice can be di(cid:27)erent for dynamic and static obstacles, as shown by Pfei(cid:27)er et al. (2017a) who use a Cartesian grid to encode the static environment and a polar grid for the location of pedestrians.”
Reinforced Imitation: Sample Efficient Deep Reinforcement Learning for Mapless Navigation by Leveraging Prior Demonstrations (2018, methodology): “The method we introduce builds upon prior work presented in [2], where a global planner is used to generate expert demonstrations in simulation.”
A Lifelong Learning Approach to Mobile Robot Navigation (2021, background): “As for physical robot navigation, learning approaches typically either imitate an expert [19], [20] or learn from trial-and-error using reinforcement learning [3], [21].”
Virtual-to-real deep reinforcement learning: Continuous control of mobile robots for mapless navigation (2017, methodology): “[14] used a deep learning model to map the laser range findings and the target position to the moving commands.”
ST-P3: End-to-end Vision-based Autonomous Driving via Spatial-Temporal Feature Learning (2022, background): “As we witness the blossom of end-to-end algorithms and success applications into various domains [38, 22], there are some attempt implementing such a philosophy in autonomous driving as well [41, 3, 15, 1, 12, 42, 53, 7, 8].”
Towards Optimally Decentralized Multi-Robot Collision Avoidance via Deep Reinforcement Learning (2017, methodology): “Sensor-level collision avoidance policies are often modeled by deep neural networks (DNNs) [10], [11] and trained using supervised learning on a large dataset.”
Distributed multi-robot collision avoidance via deep reinforcement learning for navigation in complex scenarios (2020, methodology): “To deal with unstructured environments, one data-driven motion planner is presented by Pfeiffer et al. (2017).”
Efficient Processing of Deep Neural Networks: A Tutorial and Survey (2017, methodology): “• Robotics DNNs: They have been successful in the domain of robotic tasks such as grasping with a robotic arm [32], motion planning for ground robots [33], visual navigation [4], [34], control to stabilize a quadcopter [35], and driving strategies for autonomous vehicles [36].”
GONet: A Semi-Supervised Deep Learning Approach For Traversability Estimation (2018, methodology): “Depth measurements can be obtained from 2D/3D lidars [10], [11], [12], depth cameras [13], [14], stereo camera pairs [15], or single cameras with 3D reconstruction methods [16], [17], [18].”