Collision Avoidance for Indoor Service Robots Through Multimodal Deep Reinforcement Learning

Leiva, Francisco; Lobos-Tsunekawa, Kenzo; Ruiz‐del‐Solar, Javier

doi:10.1007/978-3-030-35699-6_11

Cited by 11 publications

(8 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Kulahanek et al 26 took this idea further by removing the scene specific layers and using ideas of training with auxiliary tasks for learning better feature extraction from the image. Leiva et al 27 train a collision policy which as input combines the date from the RGB, depth image, and lasers on the robot.…”

Section: Related Workmentioning

confidence: 99%

Deep reinforcement learning for map-less goal-driven robot navigation

Dobrevski

Skočaj

2021

International Journal of Advanced Robotic Systems

View full text Add to dashboard Cite

Mobile robots that operate in real-world environments need to be able to safely navigate their surroundings. Obstacle avoidance and path planning are crucial capabilities for achieving autonomy of such systems. However, for new or dynamic environments, navigation methods that rely on an explicit map of the environment can be impractical or even impossible to use. We present a new local navigation method for steering the robot to global goals without relying on an explicit map of the environment. The proposed navigation model is trained in a deep reinforcement learning framework based on Advantage Actor–Critic method and is able to directly translate robot observations to movement commands. We evaluate and compare the proposed navigation method with standard map-based approaches on several navigation scenarios in simulation and demonstrate that our method is able to navigate the robot also without the map or when the map gets corrupted, while the standard approaches fail. We also show that our method can be directly transferred to a real robot.

show abstract

Section: Related Workmentioning

confidence: 99%

Deep reinforcement learning for map-less goal-driven robot navigation

Dobrevski

Skočaj

2021

International Journal of Advanced Robotic Systems

View full text Add to dashboard Cite

show abstract

“…Recently, great focus has been put into learning-based navigation systems [15], [14], which aim to improve their overall performance and generalize to unseen environments. Although in the machine learning community there have been several works related to visual navigation in either game-like domains [16], [10] or robotic environments [17], [18], the robotics community has been reluctant to adopt these systems, believing classic methods to perform better for generic tasks [19]. In this context, [5] shows that with sufficient training, RL-based methods can outperform classic methods on fair settings, and [20] shows that RL agents can solve visual navigation tasks with an almost perfect SPL score [15] in simulations.…”

Section: Related Work a Visual Navigationmentioning

confidence: 99%

“…However, these approaches usually use lasers as feature vectors [22] and are only valid for tasks in which the 2D observations are enough. Following the idea of using depth sensors for a seamless transfer, the use of depth cameras is logical and have been used in several works [5], [18]. However, the nature of these sensors differs greatly from the real ones (e.g., reflections, invalid values, etc.…”

Section: B Sim-to-realmentioning

confidence: 99%

Point Cloud Based Reinforcement Learning for Sim-to-Real and Partial Observability in Visual Navigation

Lobos-Tsunekawa

Harada

2020

Preprint

Self Cite

View full text Add to dashboard Cite

Reinforcement Learning (RL), among other learning-based methods, represents powerful tools to solve complex robotic tasks (e.g., actuation, manipulation, navigation, etc.), with the need for real-world data to train these systems as one of its most important limitations. The use of simulators is one way to address this issue, yet knowledge acquired in simulations does not work directly in the real-world, which is known as the sim-to-real transfer problem. While previous works focus on the nature of the images used as observations (e.g., textures and lighting), which has proven useful for a sim-to-sim transfer, they neglect other concerns regarding said observations, such as precise geometrical meanings, failing at robot-to-robot, and thus in sim-to-real transfers. We propose a method that learns on an observation space constructed by point clouds and environment randomization, generalizing among robots and simulators to achieve sim-to-real, while also addressing partial observability. We demonstrate the benefits of our methodology on the point goal navigation task, in which our method proves to be highly unaffected to unseen scenarios produced by robot-to-robot transfer, outperforms image-based baselines in robot-randomized experiments, and presents high performances in sim-to-sim conditions. Finally, we perform several experiments to validate the sim-to-real transfer to a physical domestic robot platform, confirming the out-of-the-box performance of our system.

show abstract

“…Q-Values in the algorithm are represented in tabular form which requires large memory spaces and difficult mathematical calculations. A deep reinforcement method is implemented in [8] for collision avoidance for an indoor service robot. The controller is parameterized using the neural network while DDPG is used to train the agent.…”

Section: Introductionmentioning

confidence: 99%

Application of Deep Reinforcement Learning for Tracking Control of 3WD Omnidirectional Mobile Robot

Mehmood

Shaikh²,

Ali³

2021

ITC

View full text Add to dashboard Cite

Deep reinforcement learning, the fastest growing technique, to solve real-world complex problems by creatinga simple mathematical framework. It includes an agent, action, environment, and a reward. An agent will interactwith the environment, takes an optimal action aiming to maximize the total reward. This paper proposesthe compelling technique of deep deterministic policy gradient for solving the complex continuous actionspace of 3-wheeled omnidirectional mobile robots. Three-wheeled Omnidirectional mobile robots tracking isa difficult task because of the orientation of the wheels which makes it rotate around its own axis rather tofollow the trajectory. A deep deterministic policy gradient (DDPG) algorithm has been designed to train in environmentswith continuous action space to follow the trajectory by training the neural networks defined forthe policy and value function to maximize the reward function defined for the tracking of the trajectory. DDPGagent environment is created in the Reinforcement learning toolbox in MATLAB 2019 while for Actor and criticnetwork design deep neural network designer is used. Results are shown to illustrate the effectiveness of thetechnique with a convergence of error approximately to zero.

show abstract

Collision Avoidance for Indoor Service Robots Through Multimodal Deep Reinforcement Learning

Cited by 11 publications

References 15 publications

Deep reinforcement learning for map-less goal-driven robot navigation

Deep reinforcement learning for map-less goal-driven robot navigation

Point Cloud Based Reinforcement Learning for Sim-to-Real and Partial Observability in Visual Navigation

Application of Deep Reinforcement Learning for Tracking Control of 3WD Omnidirectional Mobile Robot

Contact Info

Product

Resources

About