Ricardo Bedin Grando

Junior Costa de Jesus, Victor Augusto Kich, Alisson Henrique Kolling et al.

Reinforcement Learning (RL) has presented an impressive performance in video games through raw pixel imaging and continuous control tasks. However, RL performs poorly with high-dimensional observations such as raw pixel images. It is generally accepted that physical state-based RL policies such as laser sensor measurements give a more sample-efficient result than learning by pixels. This work presents a new approach that extracts information from a depth map estimation to teach an RL agent to perform the mapless navigation of Unmanned Aerial Vehicle (UAV). We propose the Depth-Imaged Contrastive Unsupervised Prioritized Representations in Reinforcement Learning(Depth-CUPRL) that estimates the depth of images with a prioritized replay memory. We used a combination of RL and Contrastive Learning to lead with the problem of RL based on images. From the analysis of the results with Unmanned Aerial Vehicles (UAVs), it is possible to conclude that our Depth-CUPRL approach is effective for the decision-making and outperforms state-of-the-art pixel-based approaches in the mapless navigation capability.

Victor Augusto Kich, Jair Augusto Bottega, Raul Steinmetz et al.

Kolmogorov-Arnold Networks (KANs) have shown potential as an alternative to Multi-Layer Perceptrons (MLPs) in neural networks, providing universal function approximation with fewer parameters and reduced memory usage. In this paper, we explore the use of KANs as function approximators within the Proximal Policy Optimization (PPO) algorithm. We evaluate this approach by comparing its performance to the original MLP-based PPO using the DeepMind Control Proprio Robotics benchmark. Our results indicate that the KAN-based reinforcement learning algorithm can achieve comparable performance to its MLP-based counterpart, often with fewer parameters. These findings suggest that KANs may offer a more efficient option for reinforcement learning models.

Linda Dotto de Moraes, Victor Augusto Kich, Alisson Henrique Kolling et al.

In this work, we present two Deep Reinforcement Learning (Deep-RL) approaches to enhance the problem of mapless navigation for a terrestrial mobile robot. Our methodology focus on comparing a Deep-RL technique based on the Deep Q-Network (DQN) algorithm with a second one based on the Double Deep Q-Network (DDQN) algorithm. We use 24 laser measurement samples and the relative position and angle of the agent to the target as information for our agents, which provide the actions as velocities for our robot. By using a low-dimensional sensing structure of learning, we show that it is possible to train an agent to perform navigation-related tasks and obstacle avoidance without using complex sensing information. The proposed methodology was successfully used in three distinct simulated environments. Overall, it was shown that Double Deep structures further enhance the problem for the navigation of mobile robots when compared to the ones with simple Q structures.

Victor Augusto Kich, Junior Costa de Jesus, Ricardo Bedin Grando et al.

Deep Reinforcement Learning (DRL) has produced great achievements since it was proposed, including the possibility of processing raw vision input data. However, training an agent to perform tasks based on image feedback remains a challenge. It requires the processing of large amounts of data from high-dimensional observation spaces, frame by frame, and the agent's actions are computed according to deep neural network policies, end-to-end. Image pre-processing is an effective way of reducing these high dimensional spaces, eliminating unnecessary information present in the scene, supporting the extraction of features and their representations in the agent's neural network. Modern video-games are examples of this type of challenge for DRL algorithms because of their visual complexity. In this paper, we propose a low-dimensional observation filter that allows a deep Q-network agent to successfully play in a visually complex and modern video-game, called Neon Drive.

Linda Dotto de Moraes, Victor Augusto Kich, Alisson Henrique Kolling et al.

In this study, we present two distinct approaches within the realm of Deep Reinforcement Learning (Deep-RL) aimed at enhancing mapless navigation for a ground-based mobile robot. The research methodology primarily involves a comparative analysis between a Deep-RL strategy grounded in the foundational Deep Q-Network (DQN) algorithm, and an alternative approach based on the Double Deep Q-Network (DDQN) algorithm. The agents in these approaches leverage 24 measurements from laser range sampling, coupled with the agent's positional differentials and orientation relative to the target. This amalgamation of data influences the agents' determinations regarding navigation, ultimately dictating the robot's velocities. By embracing this parsimonious sensory framework as proposed, we successfully showcase the training of an agent for proficiently executing navigation tasks and adeptly circumventing obstacles. Notably, this accomplishment is attained without a dependency on intricate sensory inputs like those inherent to image-centric methodologies. The proposed methodology is evaluated in three different real environments, revealing that Double Deep structures significantly enhance the navigation capabilities of mobile robots compared to simple Q structures.

Raul Steinmetz, Fabio Demo Rosa, Victor Augusto Kich et al.

Autonomous navigation of terrestrial robots using Reinforcement Learning (RL) from LIDAR observations remains challenging due to the high dimensionality of sensor data and the sample inefficiency of model-free approaches. Conventional policy networks struggle to process full-resolution LIDAR inputs, forcing prior works to rely on simplified observations that reduce spatial awareness and navigation robustness. This paper presents a novel model-based RL framework built on top of the DreamerV3 algorithm, integrating a Multi-Layer Perceptron Variational Autoencoder (MLP-VAE) within a world model to encode high-dimensional LIDAR readings into compact latent representations. These latent features, combined with a learned dynamics predictor, enable efficient imagination-based policy optimization. Experiments on simulated TurtleBot3 navigation tasks demonstrate that the proposed architecture achieves faster convergence and higher success rate compared to model-free baselines such as SAC, DDPG, and TD3. It is worth emphasizing that the DreamerV3-based agent attains a 100% success rate across all evaluated environments when using the full dataset of the Turtlebot3 LIDAR (360 readings), while model-free methods plateaued below 85%. These findings demonstrate that integrating predictive world models with learned latent representations enables more efficient and robust navigation from high-dimensional sensory data.

Victor Augusto Kich, Jair Augusto Bottega, Raul Steinmetz et al.

In this work, we present Curled-Dreamer, a novel reinforcement learning algorithm that integrates contrastive learning into the DreamerV3 framework to enhance performance in visual reinforcement learning tasks. By incorporating the contrastive loss from the CURL algorithm and a reconstruction loss from autoencoder, Curled-Dreamer achieves significant improvements in various DeepMind Control Suite tasks. Our extensive experiments demonstrate that Curled-Dreamer consistently outperforms state-of-the-art algorithms, achieving higher mean and median scores across a diverse set of tasks. The results indicate that the proposed approach not only accelerates learning but also enhances the robustness of the learned policies. This work highlights the potential of combining different learning paradigms to achieve superior performance in reinforcement learning applications.

Pedro Miranda Pinheiro, Armando Alves Neto, Ricardo Bedin Grando et al.

In the last decade, a great effort has been employed in the study of Hybrid Unmanned Aerial Underwater Vehicles, robots that can easily fly and dive into the water with different levels of mechanical adaptation. However, most of this literature is concentrated on physical design, practical issues of construction, and, more recently, low-level control strategies. Little has been done in the context of high-level intelligence, such as motion planning and interactions with the real world. Therefore, we proposed in this paper a trajectory planning approach that allows collision avoidance against unknown obstacles and smooth transitions between aerial and aquatic media. Our method is based on a variant of the classic Rapidly-exploring Random Tree, whose main advantages are the capability to deal with obstacles, complex nonlinear dynamics, model uncertainties, and external disturbances. The approach uses the dynamic model of the \hydrone, a hybrid vehicle proposed with high underwater performance, but we believe it can be easily generalized to other types of aerial/aquatic platforms. In the experimental section, we present simulated results in environments filled with obstacles, where the robot is commanded to perform different media movements, demonstrating the applicability of our strategy.

Ricardo Bedin Grando, Junior Costa de Jesus, Victor Augusto Kich et al.

This paper presents a novel deep reinforcement learning-based system for 3D mapless navigation for Unmanned Aerial Vehicles (UAVs). Instead of using a image-based sensing approach, we propose a simple learning system that uses only a few sparse range data from a distance sensor to train a learning agent. We based our approaches on two state-of-art double critic Deep-RL models: Twin Delayed Deep Deterministic Policy Gradient (TD3) and Soft Actor-Critic (SAC). We show that our two approaches manage to outperform an approach based on the Deep Deterministic Policy Gradient (DDPG) technique and the BUG2 algorithm. Also, our new Deep-RL structure based on Recurrent Neural Networks (RNNs) outperforms the current structure used to perform mapless navigation of mobile robots. Overall, we conclude that Deep-RL approaches based on double critic with Recurrent Neural Networks (RNNs) are better suited to perform mapless navigation and obstacle avoidance of UAVs.

Ricardo Bedin Grando, Junior Costa de Jesus, Victor Augusto Kich et al.

Since the application of Deep Q-Learning to the continuous action domain in Atari-like games, Deep Reinforcement Learning (Deep-RL) techniques for motion control have been qualitatively enhanced. Nowadays, modern Deep-RL can be successfully applied to solve a wide range of complex decision-making tasks for many types of vehicles. Based on this context, in this paper, we propose the use of Deep-RL to perform autonomous mapless navigation for Hybrid Unmanned Aerial Underwater Vehicles (HUAUVs), robots that can operate in both, air or water media. We developed two approaches, one deterministic and the other stochastic. Our system uses the relative localization of the vehicle and simple sparse range data to train the network. We compared our approaches with a traditional geometric tracking controller for mapless navigation. Based on experimental results, we can conclude that Deep-RL-based approaches can be successfully used to perform mapless navigation and obstacle avoidance for HUAUVs. Our vehicle accomplished the navigation in two scenarios, being capable to achieve the desired target through both environments, and even outperforming the geometric-based tracking controller on the obstacle-avoidance capability.