A popular framework for enforcing safe actions in Reinforcement Learning (RL) is Constrained RL, where trajectory based constraints on expected cost (or other cost measures) are employed to enforce safety and more importantly these constraints are enforced while maximizing expected reward. Most recent approaches for solving Constrained RL convert the trajectory based cost constraint into a surrogate problem that can be solved using minor modifications to RL methods. A key drawback with such approaches is an over or underestimation of the cost constraint at each state. Therefore, we provide an approach that does not modify the trajectory based cost constraint and instead imitates ``good'' trajectories and avoids ``bad'' trajectories generated from incrementally improving policies. We employ an oracle that utilizes a reward threshold (which is varied with learning) and the overall cost constraint to label trajectories as ``good'' or ``bad''. A key advantage of our approach is that we are able to work from any starting policy or set of trajectories and improve on it. In an exhaustive set of experiments, we demonstrate that our approach is able to outperform top benchmark approaches for solving Constrained RL problems, with respect to expected cost, CVaR cost, or even unknown cost constraints.

This paper presents the concepts of Artificial Intelligence, Multi-Agent-Systems, Coordination, Intelligent Robotics and Deep Reinforcement Learning. Emphasis is given on and how AI and DRL, may be efficiently used to create efficient robot skills and coordinated robotic teams, capable of performing very complex actions and tasks, such as playing a game of soccer. The paper also presents the concept of robotic soccer and the vision and structure of the RoboCup initiative with emphasis on the Humanoid Simulation 3D league and the new challenges this competition, poses. The final topics presented at the paper are based on the research developed/coordinated by the author throughout the last 22 years in the context of the FCPortugal project. The paper presents a short description of the coordination methodologies developed, such as: Strategy, Tactics, Formations, Setplays, and Coaching Languages and the use of Machine Learning to optimize the use of this concepts. The topics presented also include novel stochastic search algorithms for black box optimization and their use in the optimization of omnidirectional walking skills, robotic multi-agent learning and the creation of a humanoid kick with controlled distance. Finally, new applications using variations of the Proximal Policy Optimization algorithm and advanced modelling for robot and multi-robot learning are briefly explained with emphasis for our new humanoid sprinting and running skills and an amazing humanoid robot soccer dribbling skill. FCPortugal project enabled us to publish more than 100 papers and win several competitions in different leagues and many scientific awards at RoboCup. In total, our team won more than 40 awards in international competitions including a clear victory at the Simulation 3D League at RoboCup 2022 competition, scoring 84 goals and conceding only 2.

This paper explores the integration of Automated Guided Vehicles (AGVs) in warehouse order picking, a crucial and cost-intensive aspect of warehouse operations. The booming AGV industry, accelerated by the COVID-19 pandemic, is witnessing widespread adoption due to its efficiency, reliability, and cost-effectiveness in automating warehouse tasks. This paper focuses on enhancing the picker-to-parts system, prevalent in small to medium-sized warehouses, through the strategic use of AGVs. We discuss the benefits and applications of AGVs in various warehouse tasks, highlighting their transformative potential in improving operational efficiency. We examine the deployment of AGVs by leading companies in the industry, showcasing their varied functionalities in warehouse management. Addressing the gap in research on optimizing operational performance in hybrid environments where humans and AGVs coexist, our study delves into a dynamic picker-to-parts warehouse scenario. We propose a novel approach Neural Approximate Dynamic Programming approach for coordinating a mixed team of human and AGV workers, aiming to maximize order throughput and operational efficiency. This involves innovative solutions for non-myopic decision making, order batching, and battery management. We also discuss the integration of advanced robotics technology in automating the complete order-picking process. Through a comprehensive numerical study, our work offers valuable insights for managing a heterogeneous workforce in a hybrid warehouse setting, contributing significantly to the field of warehouse automation and logistics.

In autonomous driving, the hybrid strategy of deep reinforcement learning and cooperative adaptive cruise control (CACC) can fully utilize the advantages of the two algorithms and significantly improve the performance of car following. However, it is challenging for the traditional hybrid strategy based on fixed coefficients to adapt to mixed traffic flow scenarios, which may decrease the performance and even lead to accidents. To address the above problems, a hybrid car following strategy based on an adaptive Kalman Filter is proposed by regarding CACC and Twin Delayed Deep Deterministic Policy Gradient (TD3) algorithms. Different from traditional hybrid strategy based on fixed coefficients, the Kalman gain H, using as an adaptive coefficient, is derived from multi-timestep predictions and Monte Carlo Tree Search. At the end of study, simulation results with 4157745 timesteps indicate that, compared with the TD3 and HCFS algorithms, the proposed algorithm in this study can substantially enhance the safety of car following in mixed traffic flow without compromising the comfort and efficiency.

Generalization and sample efficiency have been long-standing issues concerning reinforcement learning, and thus the field of Offline Meta-Reinforcement Learning~(OMRL) has gained increasing attention due to its potential of solving a wide range of problems with static and limited offline data. Existing OMRL methods often assume sufficient training tasks and data coverage to apply contrastive learning to extract task representations. However, such assumptions are not applicable in several real-world applications and thus undermine the generalization ability of the representations. In this paper, we consider OMRL with two types of data limitations: limited training tasks and limited behavior diversity and propose a novel algorithm called GENTLE for learning generalizable task representations in the face of data limitations. GENTLE employs Task Auto-Encoder~(TAE), which is an encoder-decoder architecture to extract the characteristics of the tasks. Unlike existing methods, TAE is optimized solely by reconstruction of the state transition and reward, which captures the generative structure of the task models and produces generalizable representations when training tasks are limited. To alleviate the effect of limited behavior diversity, we consistently construct pseudo-transitions to align the data distribution used to train TAE with the data distribution encountered during testing. Empirically, GENTLE significantly outperforms existing OMRL methods on both in-distribution tasks and out-of-distribution tasks across both the given-context protocol and the one-shot protocol.

Deep reinforcement learning (RL) is notoriously impractical to deploy due to sample inefficiency. Meta-RL directly addresses this sample inefficiency by learning to perform few-shot learning when a distribution of related tasks is available for meta-training. While many specialized meta-RL methods have been proposed, recent work suggests that end-to-end learning in conjunction with an off-the-shelf sequential model, such as a recurrent network, is a surprisingly strong baseline. However, such claims have been controversial due to limited supporting evidence, particularly in the face of prior work establishing precisely the opposite. In this paper, we conduct an empirical investigation. While we likewise find that a recurrent network can achieve strong performance, we demonstrate that the use of hypernetworks is crucial to maximizing their potential. Surprisingly, when combined with hypernetworks, the recurrent baselines that are far simpler than existing specialized methods actually achieve the strongest performance of all methods evaluated. We provide code at https://github.com/jacooba/hyper.

In this work, we revisit the marking decisions made in the standard adaptive finite element method (AFEM). Experience shows that a na\"{i}ve marking policy leads to inefficient use of computational resources for adaptive mesh refinement (AMR). Consequently, using AFEM in practice often involves ad-hoc or time-consuming offline parameter tuning to set appropriate parameters for the marking subroutine. To address these practical concerns, we recast AMR as a Markov decision process in which refinement parameters can be selected on-the-fly at run time, without the need for pre-tuning by expert users. In this new paradigm, the refinement parameters are also chosen adaptively via a marking policy that can be optimized using methods from reinforcement learning. We use the Poisson equation to demonstrate our techniques on $h$- and $hp$-refinement benchmark problems, and our experiments suggest that superior marking policies remain undiscovered for many classical AFEM applications. Furthermore, an unexpected observation from this work is that marking policies trained on one family of PDEs are sometimes robust enough to perform well on problems far outside the training family. For illustration, we show that a simple $hp$-refinement policy trained on 2D domains with only a single re-entrant corner can be deployed on far more complicated 2D domains, and even 3D domains, without significant performance loss. For reproduction and broader adoption, we accompany this work with an open-source implementation of our methods.

In this paper, we present XuanCe, a comprehensive and unified deep reinforcement learning (DRL) library designed to be compatible with PyTorch, TensorFlow, and MindSpore. XuanCe offers a wide range of functionalities, including over 40 classical DRL and multi-agent DRL algorithms, with the flexibility to easily incorporate new algorithms and environments. It is a versatile DRL library that supports CPU, GPU, and Ascend, and can be executed on various operating systems such as Ubuntu, Windows, MacOS, and EulerOS. Extensive benchmarks conducted on popular environments including MuJoCo, Atari, and StarCraftII multi-agent challenge demonstrate the library's impressive performance.

Designing better deep networks and better reinforcement learning (RL) algorithms are both important for deep RL. This work studies the former. Specifically, the Perception and Decision-making Interleaving Transformer (PDiT) network is proposed, which cascades two Transformers in a very natural way: the perceiving one focuses on \emph{the environmental perception} by processing the observation at the patch level, whereas the deciding one pays attention to \emph{the decision-making} by conditioning on the history of the desired returns, the perceiver's outputs, and the actions. Such a network design is generally applicable to a lot of deep RL settings, e.g., both the online and offline RL algorithms under environments with either image observations, proprioception observations, or hybrid image-language observations. Extensive experiments show that PDiT can not only achieve superior performance than strong baselines in different settings but also extract explainable feature representations. Our code is available at \url{https://github.com/maohangyu/PDiT}.

In the realm of human mobility, the decision-making process for selecting the next-visit location is intricately influenced by a trade-off between spatial and temporal constraints, which are reflective of individual needs and preferences. This trade-off, however, varies across individuals, making the modeling of these spatial-temporal dynamics a formidable challenge. To address the problem, in this work, we introduce the "Spatial-temporal Induced Hierarchical Reinforcement Learning" (STI-HRL) framework, for capturing the interplay between spatial and temporal factors in human mobility decision-making. Specifically, STI-HRL employs a two-tiered decision-making process: the low-level focuses on disentangling spatial and temporal preferences using dedicated agents, while the high-level integrates these considerations to finalize the decision. To complement the hierarchical decision setting, we construct a hypergraph to organize historical data, encapsulating the multi-aspect semantics of human mobility. We propose a cross-channel hypergraph embedding module to learn the representations as the states to facilitate the decision-making cycle. Our extensive experiments on two real-world datasets validate the superiority of STI-HRL over state-of-the-art methods in predicting users' next visits across various performance metrics.