An oft-ignored challenge of real-world reinforcement learning is that the real world does not pause when agents make learning updates. As standard simulated environments do not address this real-time aspect of learning, most available implementations of RL algorithms process environment interactions and learning updates sequentially. As a consequence, when such implementations are deployed in the real world, they may make decisions based on significantly delayed observations and not act responsively. Asynchronous learning has been proposed to solve this issue, but no systematic comparison between sequential and asynchronous reinforcement learning was conducted using real-world environments. In this work, we set up two vision-based tasks with a robotic arm, implement an asynchronous learning system that extends a previous architecture, and compare sequential and asynchronous reinforcement learning across different action cycle times, sensory data dimensions, and mini-batch sizes. Our experiments show that when the time cost of learning updates increases, the action cycle time in sequential implementation could grow excessively long, while the asynchronous implementation can always maintain an appropriate action cycle time. Consequently, when learning updates are expensive, the performance of sequential learning diminishes and is outperformed by asynchronous learning by a substantial margin. Our system learns in real-time to reach and track visual targets from pixels within two hours of experience and does so directly using real robots, learning completely from scratch.
University of Southern California（南加州大学）；Google Brain（谷歌大脑）
In principle, reinforcement learning and policy search methods can enable robots to learn highly complex and general skills that may allow them to function amid the complexity and diversity of the real world. However, training a policy that generalizes well across a wide range of real-world conditions requires far greater quantity and diversity of experience than is practical to collect with a single robot. Fortunately, it is possible for multiple robots to share their experience with one another, and thereby, learn a policy collectively. In this work, we explore distributed and asynchronous p…
University of Cambridge（剑桥大学）；Google DeepMind（谷歌DeepMind）；UC Berkeley（加州大学伯克利分校）
Reinforcement learning holds the promise of enabling autonomous robots to learn large repertoires of behavioral skills with minimal human intervention. However, robotic applications of reinforcement learning often compromise the autonomy of the learning process in favor of achieving training times that are practical for real physical systems. This typically involves introducing hand-engineered policy representations and human-supplied demonstrations. Deep reinforcement learning alleviates this limitation by training general-purpose neural network policies, but applications of direct deep reinf…
Significant progress has been made in the area of model-based reinforcement learning. State-of-the-art algorithms are now able to match the asymptotic performance of model-free methods while being significantly more data efficient. However, this success has come at a price: state-of-the-art model-based methods require significant computation interleaved with data collection, resulting in run times that take days, even if the amount of agent interaction might be just hours or even minutes. When considering the goal of learning in real-time on real robots, this means these state-of-the-art model…
Southern University of Science and Technology（南方科技大学）；NetEase Games AI Lab（网易游戏人工智能实验室）
Despite the emerging progress of integrating evolutionary computation into reinforcement learning, the absence of a high-performance platform endowing composability and massive parallelism causes non-trivial difficulties for research and applications related to asynchronous commercial games. Here we introduce Lamarckian - an open-source platform featuring support for evolutionary reinforcement learning scalable to distributed computing resources. To improve the training speed and data efficiency, Lamarckian adopts optimized communication methods and an asynchronous evolutionary reinforcement l…
The relationship between a reinforcement learning (RL) agent and an asynchronous environment is often ignored. Frequently used models of the interaction between an agent and its environment, such as Markov Decision Processes (MDP) or Semi-Markov Decision Processes (SMDP), do not capture the fact that, in an asynchronous environment, the state of the environment may change during computation performed by the agent. In an asynchronous environment, minimizing reaction time---the time it takes for an agent to react to an observation---also minimizes the time in which the state of the environment m…
The state-of-the-art multi-agent reinforcement learning (MARL) methods have provided promising solutions to a variety of complex problems. Yet, these methods all assume that agents perform synchronized primitive-action executions so that they are not genuinely scalable to long-horizon real-world multi-agent/robot tasks that inherently require agents/robots to asynchronously reason about high-level action selection at varying time durations. The Macro-Action Decentralized Partially Observable Markov Decision Process (MacDec-POMDP) is a general formalization for asynchronous decision-making unde…
Deep reinforcement learning, and especially the Asynchronous Advantage Actor-Critic algorithm, has been successfully used to achieve super-human performance in a variety of video games. Starcraft II is a new challenge for the reinforcement learning community with the release of pysc2 learning environment proposed by Google Deepmind and Blizzard Entertainment. Despite being a target for several AI developers, few have achieved human level performance. In this project we explain the complexities of this environment and discuss the results from our experiments on the environment. We have compared…
The development of quantum machine learning (QML) has received a lot of interest recently thanks to developments in both quantum computing (QC) and machine learning (ML). One of the ML paradigms that can be utilized to address challenging sequential decision-making issues is reinforcement learning (RL). It has been demonstrated that classical RL can successfully complete many difficult tasks. A leading method of building quantum RL agents relies on the variational quantum circuits (VQC). However, training QRL algorithms with VQCs requires significant amount of computational resources. This iss…
Unmanned aerial vehicles (UAVs) have been widely used in military warfare. In this paper, we formulate the autonomous motion control (AMC) problem as a Markov decision process (MDP) and propose an advanced deep reinforcement learning (DRL) method that allows UAVs to execute complex tasks in large-scale dynamic three-dimensional (3D) environments. To overcome the limitations of the prioritized experience replay (PER) algorithm and improve performance, the proposed asynchronous curriculum experience replay (ACER) uses multithreads to asynchronously update the priorities, assigns the true priorit…
This article surveys reinforcement learning approaches in social robotics. Reinforcement learning is a framework for decision-making problems in which an agent interacts through trial-and-error with its environment to discover an optimal behavior. Since interaction is a key component in both reinforcement learning and social robotics, it can be a well-suited approach for real-world interactions with physically embodied social robots. The scope of the paper is focused particularly on studies that include social physical robots and real-world human-robot interactions with users. We present a tho…
Reinforcement learning is a promising method to accomplish robotic control tasks. The task of playing musical instruments is, however, largely unexplored because it involves the challenge of achieving sequential goals - melodies - that have a temporal dimension. In this paper, we address robotic musicianship by introducing a temporal extension to goal-conditioned reinforcement learning: Time-dependent goals. We demonstrate that these can be used to train a robotic musician to play the theremin instrument. We train the robotic agent in simulation and transfer the acquired policy to a real-world…
Shanghai Jiao Tong University（上海交通大学）；Ruijin Hospital affiliated to Shanghai Jiao Tong University School of Medicine（上海交通大学医学院附属瑞金医院）；Shanghai Ruijin Rehabilitation Hospital（上海瑞金康复医院）
Compared to rigid robots that are generally studied in reinforcement learning, the physical characteristics of some sophisticated robots such as soft or continuum robots are higher complicated. Moreover, recent reinforcement learning methods are data-inefficient and can not be directly deployed to the robot without simulation. In this paper, we propose an efficient reinforcement learning method based on inexplicit prior knowledge in response to such problems. We first corroborate the method by simulation and employed directly in the real world. By using our method, we can achieve active visual…
One of the biggest hurdles robotics faces is the facet of sophisticated and hard-to-engineer behaviors. Reinforcement learning offers a set of tools, and a framework to address this problem. In parallel, the misgivings of robotics offer a solid testing ground and evaluation metric for advancements in reinforcement learning. The two disciplines go hand-in-hand, much like the fields of Mathematics and Physics. By means of this survey paper, we aim to invigorate links between the research communities of the two disciplines by focusing on the work done in reinforcement learning for locomotive and …
Fraunhofer Institute for Integrated Circuits IIS（弗劳恩霍夫集成电路研究所 IIS）；Friedrich-Alexander-University Erlangen-Nuremberg (FAU)（弗里德里希-亚历山大-埃尔兰根-纽伦堡大学（FAU））
This paper presents a novel motion and trajectory planning algorithm for nonholonomic mobile robots that uses recent advances in deep reinforcement learning. Starting from a random initial state, i.e., position, velocity and orientation, the robot reaches an arbitrary target state while taking both kinematic and dynamic constraints into account. Our deep reinforcement learning agent not only processes a continuous state space it also executes continuous actions, i.e., the acceleration of wheels and the adaptation of the steering angle. We evaluate our motion and trajectory planning on a mobile…
University of Southern California（南加州大学）；Intel Labs（英特尔实验室）
Increasing the scale of reinforcement learning experiments has allowed researchers to achieve unprecedented results in both training sophisticated agents for video games, and in sim-to-real transfer for robotics. Typically such experiments rely on large distributed systems and require expensive hardware setups, limiting wider access to this exciting area of research. In this work we aim to solve this problem by optimizing the efficiency and resource utilization of reinforcement learning algorithms instead of relying on distributed computation. We present the "Sample Factory", a high-throughput…
Khoury College of Computer Sciences（科里计算机科学学院）；Northeastern University（东北大学）
In real-world multi-robot systems, performing high-quality, collaborative behaviors requires robots to asynchronously reason about high-level action selection at varying time durations. Macro-Action Decentralized Partially Observable Markov Decision Processes (MacDec-POMDPs) provide a general framework for asynchronous decision making under uncertainty in fully cooperative multi-agent tasks. However, multi-agent deep reinforcement learning methods have only been developed for (synchronous) primitive-action problems. This paper proposes two Deep Q-Network (DQN) based methods for learning decent…
The asynchronous nature of the state-of-the-art reinforcement learning algorithms such as the Asynchronous Advantage Actor-Critic algorithm, makes them exceptionally suitable for CPU computations. However, given the fact that deep reinforcement learning often deals with interpreting visual information, a large part of the train and inference time is spent performing convolutions. In this work we present our results on learning strategies in Atari games using a Convolutional Neural Network, the Math Kernel Library and TensorFlow 0.11rc0 machine learning framework. We also analyze effects of asy…
South Westphalia University of Applied Sciences（西南法伦应用科学大学）
This paper introduces a novel combination of scheduling control on a flexible robot manufacturing cell with curiosity based reinforcement learning. Reinforcement learning has proved to be highly successful in solving tasks like robotics and scheduling. But this requires hand tuning of rewards in problem domains like robotics and scheduling even where the solution is not obvious. To this end, we apply a curiosity based reinforcement learning, using intrinsic motivation as a form of reward, on a flexible robot manufacturing cell to alleviate this problem. Further, the learning agents are embedde…
The University of Tokyo（东京大学）；Keio University（庆应义塾大学）
We demonstrate the first reinforcement-learning application for robots equipped with an event camera. Because of the considerably lower latency of the event camera, it is possible to achieve much faster control of robots compared with the existing vision-based reinforcement-learning applications using standard cameras. To handle a stream of events for reinforcement learning, we introduced an image-like feature and demonstrated the feasibility of training an agent in a simulator for two tasks: fast collision avoidance and obstacle tracking. Finally, we set up a robot with an event camera in the…
Korea Advanced Institute of Science and Technology (KAIST)（韩国科学技术院）
Deep reinforcement learning (DRL) algorithms and evolution strategies (ES) have been applied to various tasks, showing excellent performances. These have the opposite properties, with DRL having good sample efficiency and poor stability, while ES being vice versa. Recently, there have been attempts to combine these algorithms, but these methods fully rely on synchronous update scheme, making it not ideal to maximize the benefits of the parallelism in ES. To solve this challenge, asynchronous update scheme was introduced, which is capable of good time-efficiency and diverse policy exploration. …
Reinforcement learning (RL) is a promising field to enhance robotic autonomy and decision making capabilities for space robotics, something which is challenging with traditional techniques due to stochasticity and uncertainty within the environment. RL can be used to enable lunar cave exploration with infrequent human feedback, faster and safer lunar surface locomotion or the coordination and collaboration of multi-robot systems. However, there are many hurdles making research challenging for space robotic applications using RL and machine learning, particularly due to insufficient resources f…
University College Dublin（都柏林大学学院）；Resero Ltd（Resero有限公司）；Universidad de León（莱昂大学）
Reinforcement learning has shown great promise in robotics thanks to its ability to develop efficient robotic control procedures through self-training. In particular, reinforcement learning has been successfully applied to solving the reaching task with robotic arms. In this paper, we define a robust, reproducible and systematic experimental procedure to compare the performance of various model-free algorithms at solving this task. The policies are trained in simulation and are then transferred to a physical robotic manipulator. It is shown that augmenting the reward signal with the Hindsight …
Intel（英特尔）；Biz On Sp. z o.o.（Biz On 有限责任公司）；Polish Academy of Sciences（波兰科学院）
We present a study in Distributed Deep Reinforcement Learning (DDRL) focused on scalability of a state-of-the-art Deep Reinforcement Learning algorithm known as Batch Asynchronous Advantage ActorCritic (BA3C). We show that using the Adam optimization algorithm with a batch size of up to 2048 is a viable choice for carrying out large scale machine learning computations. This, combined with careful reexamination of the optimizer's hyperparameters, using synchronous training on the node level (while keeping the local, single node part of the algorithm asynchronous) and minimizing the memory footp…
Asynchronous Distributed Reinforcement Learning for LQR Control via Zeroth-Order Block Coordinate Descent
Chongqing University（重庆大学）；North Carolina State University（北卡罗来纳州立大学）；Oklahoma State University（俄克拉荷马州立大学）；DEVCOM Army Research Laboratory（DEVCOM陆军研究实验室）
Reinforcement learning is a promising approach to developing hard-to-engineer adaptive solutions for complex and diverse robotic tasks. However, learning with real-world robots is often unreliable and difficult, which resulted in their low adoption in reinforcement learning research. This difficulty is worsened by the lack of guidelines for setting up learning tasks with robots. In this work, we develop a learning task with a UR5 robotic arm to bring to light some key elements of a task setup and study their contributions to the challenges with robots. We find that learning performance can be …
We introduce SLM Lab, a software framework for reproducible reinforcement learning (RL) research. SLM Lab implements a number of popular RL algorithms, provides synchronous and asynchronous parallel experiment execution, hyperparameter search, and result analysis. RL algorithms in SLM Lab are implemented in a modular way such that differences in algorithm performance can be confidently ascribed to differences between algorithms, not between implementations. In this work we present the design choices behind SLM Lab and use it to produce a comprehensive single-codebase RL algorithm benchmark. In…
Facebook AI Research（Facebook人工智能研究院）；McGill University（麦吉尔大学）
Multi-simulator training has contributed to the recent success of Deep Reinforcement Learning by stabilizing learning and allowing for higher training throughputs. We propose Gossip-based Actor-Learner Architectures (GALA) where several actor-learners (such as A2C agents) are organized in a peer-to-peer communication topology, and exchange information through asynchronous gossip in order to take advantage of a large number of distributed simulators. We prove that GALA agents remain within an epsilon-ball of one-another during training when using loosely coupled asynchronous communication. By r…
Colorado School of Mines（科罗拉多矿业学院）；Toyota Motor North America（丰田汽车北美公司）
Collaborative localization is an essential capability for a team of robots such as connected vehicles to collaboratively estimate object locations from multiple perspectives with reliant cooperation. To enable collaborative localization, four key challenges must be addressed, including modeling complex relationships between observed objects, fusing observations from an arbitrary number of collaborating robots, quantifying localization uncertainty, and addressing latency of robot communications. In this paper, we introduce a novel approach that integrates uncertainty-aware spatiotemporal graph …
Université Pierre et Marie Curie - Paris 6（巴黎第六大学皮埃尔与玛丽·居里大学）；LIP6-CNRS（巴黎第六大学计算机实验室（法国国家科学研究中心））
We propose the first deterministic algorithm that tolerates up to $f$ byzantine faults in $3f+1$-sized networks and performs in the asynchronous CORDA model. Our solution matches the previously established lower bound for the semi-synchronous ATOM model on the number of tolerated Byzantine robots. Our algorithm works under bounded scheduling assumptions for oblivious robots moving in a uni-dimensional space.
Khoury College of Computer Sciences（库利计算机科学学院）；Northeastern University（东北大学）
Synchronizing decisions across multiple agents in realistic settings is problematic since it requires agents to wait for other agents to terminate and communicate about termination reliably. Ideally, agents should learn and execute asynchronously instead. Such asynchronous methods also allow temporally extended actions that can take different amounts of time based on the situation and action executed. Unfortunately, current policy gradient methods are not applicable in asynchronous settings, as they assume that agents synchronously reason about action selection at every time step. To allow asy…
Université Pierre et Marie Curie - Paris 6（巴黎第六大学（皮埃尔和玛丽·居里大学））；Ben Gurion University of the Negev（内盖夫本·古里安大学）
This paper introduces the \emph{RoboCast} communication abstraction. The RoboCast allows a swarm of non oblivious, anonymous robots that are only endowed with visibility sensors and do not share a common coordinate system, to asynchronously exchange information. We propose a generic framework that covers a large class of asynchronous communication algorithms and show how our framework can be used to implement fundamental building blocks in robot networks such as gathering or stigmergy. In more details, we propose a RoboCast algorithm that allows robots to broadcast their local coordinate syste…
Decentralized swarm robotic solutions to searching for targets that emit a spatially varying signal promise task parallelism, time efficiency, and fault tolerance. It is, however, challenging for swarm algorithms to offer scalability and efficiency, while preserving mathematical insights into the exhibited behavior. A new decentralized search method (called Bayes-Swarm), founded on batch Bayesian Optimization (BO) principles, is presented here to address these challenges. Unlike swarm heuristics approaches, Bayes-Swarm decouples the knowledge generation and task planning process, thus preservi…
Jadavpur University（贾达普大学）；University of Washington（华盛顿大学）；Indian Statistical Institute（印度统计研究所）
This paper presents a deterministic algorithm for forming a given asymmetric pattern in finite time by a set of autonomous, homogeneous, oblivious mobile robots under the CORDA model. The robots are represented as points on the 2D plane. There is no explicit communication between the robots. The robots coordinate among themselves by observing the positions of the other robots on the plane. Initially all the robots are assumed to be stationary. The robots have local coordinate systems defined by Sense of Direction (SoD), orientation or chirality and scale. Initially the robots are in asymmetric…
Indian Statistical Institute（印度统计研究所）；Jadavpur University（贾达普大学）
The traditional distributed model of autonomous, homogeneous, mobile point robots usually assumes that the robots do not create any visual obstruction for the other robots, i.e., the robots are see through. In this paper, we consider a slightly more realistic model, by incorporating the notion of obstructed visibility (i.e., robots are not see through) for other robots. Under the new model of visibility, a robot may not have the full view of its surroundings. Many of the existing algorithms demand that each robot should have the complete knowledge of the positions of other robots. Since, visio…
In this paper, we study the Online Network Formation Problem (ONFP) for a mobile multi-robot system. Consider a group of robots with a bounded communication range operating in a large open area. One of the robots has a piece of information which has to be propagated to all other robots. What strategy should the robots pursue to disseminate the information to the rest of the robots as quickly as possible? The initial locations of the robots are unknown to each other, therefore the problem must be solved in an online fashion. For this problem, we present an algorithm whose competitive ratio is $…
