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

Dobrevski, Matej; Skočaj, Danijel

doi:10.1177/1729881421992621

Cited by 24 publications

(14 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Tai et al [7] merged range findings, position relative to the target and previous velocity as input and complete collision-free navigation with DDPG. Matej et al [8] grouped the range scanner into 30 bins as input and trained the agent with A2C [14] and recurrent long short-term memory [15] (LSTM) neural network. In this paper, we propose an effective target-driven framework leveraging multiple information as input for collision-free mapless navigation.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Target-driven multi-input mapless robot navigation with deep reinforcement learning

Hua

2023

J. Phys.: Conf. Ser.

View full text Add to dashboard Cite

Collision-free navigation without pre-built maps is a promising and challenging problem. To make full use of the information in the environment, we propose a multi-input mapless navigation framework that combines both vision-based and ranging-based approaches. It takes the position and orientation of the robot relative to the target as the reference, incorporating RGB-D images and distance measurements as state inputs to the deep reinforcement learning model. In the framework, Dual-VAE is designed to encode RGB images and depth images simultaneously, in order to effectively mine the intrinsic information of RGB-D images. Moreover, we engineer a special reward function that enables the robot to navigate with a smooth trajectory and leave redundant space with obstacles for other tasks. Target-driven multi-input mapless navigation framework has been experimentally proven to achieve satisfactory performance in simulation environments. Our source code and simulation environments are released at https://github.com/YosephYu/MultiInput-Mapless-Navigation.

show abstract

Section: Related Workmentioning

confidence: 99%

“…With the rapid progress of deep reinforcement learning, it makes mapless navigation feasible. In the past few years, many scholars have implemented vision-based [4,5,6] or ranging-based [7,8] mapless navigation.…”

Section: Introductionmentioning

confidence: 99%

Target-driven multi-input mapless robot navigation with deep reinforcement learning

Hua

2023

J. Phys.: Conf. Ser.

View full text Add to dashboard Cite

show abstract

“…Deep Reinforcement Learning (DRL) has been widely used in complex control problems under uncertainties. Dobrevksy et al [23] proposed a navigation model based on the advantage actor-critic method, which can directly project robot observations to the motion commands. Risk-aware navigation that learns to perform conservative actions when there are high chances of collision and moves faster otherwise has been addressed by Kahn et al [21].…”

Section: Learning-based Navigationmentioning

confidence: 99%

AdaptiveON: Adaptive Outdoor Local Navigation Method For Stable and Reliable Actions

Ji¹,

Weerakoon²,

Guan³

et al. 2022

Preprint

View full text Add to dashboard Cite

We present a novel outdoor navigation algorithm to generate stable and efficient actions to navigate a robot to the goal. We use a multi-stage training pipeline and show that our model produces policies that result in stable and reliable robot navigation on complex terrains. Based on the Proximal Policy Optimization (PPO) algorithm, we developed a novel method to achieve multiple capabilities for outdoor navigation tasks, namely: alleviating the robot's drifting, keeping the robot stable on bumpy terrains, avoiding climbing on hills with steep elevation changes, and collision avoidance. Our training process mitigates the reality(sim-to-real) gap by introducing more generalized environmental and robotic parameters and training with rich features of Lidar perception in the Unity simulator. We evaluate our method in both simulation and the real world with Clearpath Husky and Jackal. Additionally, we compare our method against the state-of-the-art approaches and show that in the real world it improves stability by at least 30.7% on uneven terrains, reduces drifting by 8.08%, and for high hills our trained policy keeps small changes of the elevation of the robot at each motion step by preventing the robot from moving on areas with high gradients.

show abstract

“…DRL has spurred a lot of significant breakthroughs in many applications, such as manipulator controlling [18], autonomous driving [19,20], and games [21,22]. Considering the strengths of DRL, researchers attempt to apply it to tackle the PointGoal navigation problem [4,[23][24][25][26]. Compared with the previous traditional methods, the DRL-based methods avoid extensive hand-engineering but only need to learn the complete navigation system directly from the data.…”

Section: Introductionmentioning

confidence: 99%

“…In summary, approaches to solve PointGoal navigation can broadly be classified into two categories: (1) traditional navigation methods that decompose navigation into localization, mapping and planning [12][13][14][15] or (2) learning neural policies using DRL [4,[23][24][25][26][27][28][29]35]. Traditional navigation methods have the weakness of intensive computational demand and involve numerous parameters that need to be tuned manually.…”

Section: Introductionmentioning

confidence: 99%

Safe-Nav: learning to prevent PointGoal navigation failure in unknown environments

Jin

Meng

Dai

et al. 2022

Complex Intell. Syst.

View full text Add to dashboard Cite

Training robots to safely navigate (Safe-Nav) in uncertain complex environments using the RGB-D sensor is quite challenging as it involves the performance of different tasks such as obstacle avoidance, optimal path planning, and control. Traditional navigation approaches cannot generate suitable paths which guarantee enough visible features. Recent learning-based methods are still not mature enough due to their proneness to collisions and prohibitive computational cost. This paper focuses on generating safe trajectories to the desired goal while avoiding collisions and tracking failure in unknown complex environments. We present Safe-Nav, a hierarchical framework composed of the visual simultaneous localization and mapping (SLAM) module, the global planner module and the local planner module. The visual SLAM module generates the navigation map and the robot pose. The global planner module plans a local waypoint on the real-time navigation map. In the local planner module, a deep-reinforcement-learning-based (DRL-based) policy is presented for taking safe actions towards local waypoints. Our DRL-based policy can learn different navigation skills (e.g., avoiding collisions and avoiding tracking failure) through specialized modes without any supervisory signals when the PointGoal-navigation-specied reward is provided. We have demonstrated the performance of our proposed Safe-Nav in the Habitat simulation environment. Our approach outperforms the recent learning-based method and conventional navigation approach with relative improvements of over 205% (0.55 vs. 0.18) and 139% (0.55 vs. 0.23) in the success rate, respectively.

show abstract

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

Cited by 24 publications

References 20 publications

Target-driven multi-input mapless robot navigation with deep reinforcement learning

Target-driven multi-input mapless robot navigation with deep reinforcement learning

AdaptiveON: Adaptive Outdoor Local Navigation Method For Stable and Reliable Actions

Safe-Nav: learning to prevent PointGoal navigation failure in unknown environments

Contact Info

Product

Resources

About