Transferring reinforcement learning policies trained in physics simulation to the real hardware remains a challenge, known as the "sim-to-real" gap. Domain randomization is a simple yet effective technique to address dynamics discrepancies across source and target domains, but its success generally depends on heuristics and trial-and-error. In this work we investigate the impact of randomized parameter selection on policy transferability across different types of domain discrepancies. Contrary to common practice in which kinematic parameters are carefully measured while dynamic parameters are randomized, we found that virtually randomizing kinematic parameters (e.g., link lengths) during training in simulation generally outperforms dynamic randomization. Based on this finding, we introduce a new domain adaptation algorithm that utilizes simulated kinematic parameters variation. Our algorithm, Multi-Policy Bayesian Optimization, trains an ensemble of universal policies conditioned on virtual kinematic parameters and efficiently adapts to the target environment using a limited number of target domain rollouts. We showcase our findings on a simulated quadruped robot in five different target environments covering different aspects of domain discrepancies.
University of British Columbia（不列颠哥伦比亚大学）；NVIDIA（英伟达）；University of Toronto（多伦多大学）；Vector Institute（向量研究院）
Understanding the gap between simulation and reality is critical for reinforcement learning with legged robots, which are largely trained in simulation. However, recent work has resulted in sometimes conflicting conclusions with regard to which factors are important for success, including the role of dynamics randomization. In this paper, we aim to provide clarity and understanding on the role of dynamics randomization in learning robust locomotion policies for the Laikago quadruped robot. Surprisingly, in contrast to prior work with the same robot model, we find that direct sim-to-real transf…
Safe Continual Domain Adaptation after Sim2Real Transfer of Reinforcement Learning Policies in Robotics
Technical University of Munich（慕尼黑工业大学）；University of California, Berkeley（加州大学伯克利分校）；Technische Universität Berlin（柏林工业大学）；L3S Research Center, Leibniz Universität Hannover（莱布尼茨汉诺威大学L3S研究中心）
Real world data, especially in the domain of robotics, is notoriously costly to collect. One way to circumvent this can be to leverage the power of simulation to produce large amounts of labelled data. However, training models on simulated images does not readily transfer to real-world ones. Using domain adaptation methods to cross this "reality gap" requires a large amount of unlabelled real-world data, whilst domain randomization alone can waste modeling power. In this paper, we present Randomized-to-Canonical Adaptation Networks (RCANs), a novel approach to crossing the visual reality gap t…
Training deep reinforcement learning (DRL) locomotion policies often require massive amounts of data to converge to the desired behaviour. In this regard, simulators provide a cheap and abundant source. For successful sim-to-real transfer, exhaustively engineered approaches such as system identification, dynamics randomization, and domain adaptation are generally employed. As an alternative, we investigate a simple strategy of random force injection (RFI) to perturb system dynamics during training. We show that the application of random forces enables us to emulate dynamics randomization. This…
Mind the Gap: Towards Generalizable Autonomous Penetration Testing via Domain Randomization and Meta-Reinforcement Learning
College of Electronic Engineering, National University of Defense Technology（国防科技大学电子工程学院）；Beijing National Research Center for Information Science and Technology, Tsinghua University（清华大学北京信息科学与技术国家研究中心）；College of Computer Science and Technology, National University of Defense Technology（国防科技大学计算机学院）
Reinforcement learning has shown a wide usage in robotics tasks, such as insertion and grasping. However, without a practical sim2real strategy, the policy trained in simulation could fail on the real task. There are also wide researches in the sim2real strategies, but most of those methods rely on heavy image rendering, domain randomization training, or tuning. In this work, we solve the insertion task using a pure visual reinforcement learning solution with minimum infrastructure requirement. We also propose a novel sim2real strategy, Real2Sim, which provides a novel and easier solution in p…
Université de Montréal（蒙特利尔大学）；Mila（米拉）；Polytechnique Montréal（蒙特利尔综合理工学院）；CIFAR AI Chair（加拿大高级研究院人工智能主席）
Gradient-based meta-learners such as Model-Agnostic Meta-Learning (MAML) have shown strong few-shot performance in supervised and reinforcement learning settings. However, specifically in the case of meta-reinforcement learning (meta-RL), we can show that gradient-based meta-learners are sensitive to task distributions. With the wrong curriculum, agents suffer the effects of meta-overfitting, shallow adaptation, and adaptation instability. In this work, we begin by highlighting intriguing failure cases of gradient-based meta-RL and show that task distributions can wildly affect algorithmic out…
ABV-Indian Institute of Information Technology and Management, Gwalior（ABV印度信息技术与管理学院（瓜廖尔））
Robotics has proved to be an indispensable tool in many industrial as well as social applications, such as warehouse automation, manufacturing, disaster robotics, etc. In most of these scenarios, damage to the agent while accomplishing mission-critical tasks can result in failure. To enable robotic adaptation in such situations, the agent needs to adopt policies which are robust to a diverse set of damages and must do so with minimum computational complexity. We thus propose a damage aware control architecture which diagnoses the damage prior to gait selection while also incorporating domain r…
IRI, CSIC-UPC（加泰罗尼亚理工大学西班牙国家研究委员会机器人研究所）；UC Berkeley（加州大学伯克利分校）；NYU（纽约大学）；UC San Diego（加州大学圣迭戈分校）；Technical University of Denmark（丹麦技术大学）
In most real world scenarios, a policy trained by reinforcement learning in one environment needs to be deployed in another, potentially quite different environment. However, generalization across different environments is known to be hard. A natural solution would be to keep training after deployment in the new environment, but this cannot be done if the new environment offers no reward signal. Our work explores the use of self-supervision to allow the policy to continue training after deployment without using any rewards. While previous methods explicitly anticipate changes in the new enviro…
CARO: Contact-Agnostic Residual Observation for Zero-Shot Robust Quadruped Locomotion
School of Aeronautic Science and Engineering, Beihang University（北京航空航天大学航空科学与工程学院）；School of Automation Science and Electrical Engineering, Beihang University（北京航空航天大学自动化科学与电气工程学院）
Simulation-to-simulation and simulation-to-real world transfer of neural network models have been a difficult problem. To close the reality gap, prior methods to simulation-to-real world transfer focused on domain adaptation, decoupling perception and dynamics and solving each problem separately, and randomization of agent parameters and environment conditions to expose the learning agent to a variety of conditions. While these methods provide acceptable performance, the computational complexity required to capture a large variation of parameters for comprehensive scenarios on a given task suc…
Transferable Reinforcement Learning via Probabilistic Latent Embeddings and Dynamic Policy Adaptation for Sim-to-Real Deployment
Lyles School of Civil and Construction Engineering, Purdue University, West Lafayette, USA（普渡大学土木与建设工程学院）；Elmore Family School of Electrical and Computer Engineering, Purdue University, West Lafayette, USA（普渡大学埃尔莫尔家庭电气与计算机工程学院）
Collecting and automatically obtaining reward signals from real robotic visual data for the purposes of training reinforcement learning algorithms can be quite challenging and time-consuming. Methods for utilizing unlabeled data can have a huge potential to further accelerate robotic learning. We consider here the problem of performing manipulation tasks from pixels. In such tasks, choosing an appropriate state representation is crucial for planning and control. This is even more relevant with real images where noise, occlusions and resolution affect the accuracy and reliability of state estim…
Institute for Systems and Robotics（系统与机器人研究所）；Instituto Superior Técnico（里斯本高等理工学院）；Universidade de Lisboa（里斯本大学）
Humanoid robots have complex bodies and kinematic chains with several Degrees-of-Freedom (DoF) which are difficult to model. Learning the parameters of a kinematic model can be achieved by observing the position of the robot links during prospective motions and minimising the prediction errors. This work proposes a movement efficient approach for estimating online the body-schema of a humanoid robot arm in the form of Denavit-Hartenberg (DH) parameters. A cost-sensitive active learning approach based on the A-Optimality criterion is used to select optimal joint configurations. The chosen joint…
Volkswagen AG（大众汽车公司）；Technische Universität Darmstadt（达姆施塔特工业大学）
General-purpose planning algorithms for automated driving combine mission, behavior, and local motion planning. Such planning algorithms map features of the environment and driving kinematics into complex reward functions. To achieve this, planning experts often rely on linear reward functions. The specification and tuning of these reward functions is a tedious process and requires significant experience. Moreover, a manually designed linear reward function does not generalize across different driving situations. In this work, we propose a deep learning approach based on inverse reinforcement …
Stevens Institute of Technology（斯蒂文斯理工学院）；Massachusetts Institute of Technology（麻省理工学院）；Wing, Alphabet Inc.（Alphabet公司Wing部门）
This paper studies the problem of autonomous exploration under localization uncertainty for a mobile robot with 3D range sensing. We present a framework for self-learning a high-performance exploration policy in a single simulation environment, and transferring it to other environments, which may be physical or virtual. Recent work in transfer learning achieves encouraging performance by domain adaptation and domain randomization to expose an agent to scenarios that fill the inherent gaps in sim2sim and sim2real approaches. However, it is inefficient to train an agent in environments with rand…
Deep reinforcement learning has recently seen huge success across multiple areas in the robotics domain. Owing to the limitations of gathering real-world data, i.e., sample inefficiency and the cost of collecting it, simulation environments are utilized for training the different agents. This not only aids in providing a potentially infinite data source, but also alleviates safety concerns with real robots. Nonetheless, the gap between the simulated and real worlds degrades the performance of the policies once the models are transferred into real robots. Multiple research efforts are therefore…
Deep reinforcement learning (DRL) is a promising approach to solve complex control tasks by learning policies through interactions with the environment. However, the training of DRL policies requires large amounts of training experiences, making it impractical to learn the policy directly on physical systems. Sim-to-real approaches leverage simulations to pretrain DRL policies and then deploy them in the real world. Unfortunately, the direct real-world deployment of pretrained policies usually suffers from performance deterioration due to the different dynamics, known as the reality gap. Recen…
Synergies have been adopted in prosthetic limb applications to reduce complexity of design, but typically involve a single synergy setting for a population and ignore individual preference or adaptation capacity. However, personalization of the synergy setting is necessary for the effective operation of the prosthetic device. Two major challenges hinder the personalization of synergies in human-prosthesis interfaces. The first is related to the process of human motor adaptation and the second to the variation in motor learning dynamics of individuals. In this paper, a systematic personalizatio…
Munich Institute of Robotics and Machine Intelligence（慕尼黑机器人与机器智能研究所）；Technische Universitat Munchen（慕尼黑工业大学）
Manipulability ellipsoids efficiently capture the human pose and reveal information about the task at hand. Their use in task-dependent robot teaching - particularly their transfer from a teacher to a learner - can advance emulation of human-like motion. Although in recent literature focus is shifted towards manipulability transfer between two robots, the adaptation to the capabilities of the other kinematic system is to date not addressed and research in transfer from human to robot is still in its infancy. This work presents a novel manipulability domain adaptation method for the transfer of…
Rapid Embodiment Adaptation for Quadrupedal Locomotion
UC San Diego（加州大学圣迭戈分校）；Stanford University（斯坦福大学）；TU Darmstadt（达姆施塔特工业大学）；Sudo AI GmbH（Sudo AI 有限公司）；German Research Center for AI (DFKI)（德国人工智能研究中心）；Robotics Institute Germany（德国机器人研究所）
Providing an accurate and efficient assessment of operative difficulty is important for designing robot-assisted teleoperation interfaces that are easy and natural for human operators to use. In this paper, we aim to develop a data-driven approach to numerically characterize the operative difficulty demand of complex teleoperation. In effort to provide an entirely task-independent assessment, we consider using only data collected from the human user including: (1) physiological response, and (2) movement kinematics. By leveraging an unsupervised domain adaptation technique, our approach learns…
The Chinese University of Hong Kong（香港中文大学）；The Hong Kong Polytechnic University（香港理工大学）
Automated surgical gesture recognition is of great importance in robot-assisted minimally invasive surgery. However, existing methods assume that training and testing data are from the same domain, which suffers from severe performance degradation when a domain gap exists, such as the simulator and real robot. In this paper, we propose a novel unsupervised domain adaptation framework which can simultaneously transfer multi-modality knowledge, i.e., both kinematic and visual data, from simulator to real robot. It remedies the domain gap with enhanced transferable features by using temporal cues…
Mounting evidence suggests that adaptation is a crucial mechanism for rehabilitation robots in promoting motor learning. Yet, it is commonly based on robot-derived movement kinematics, which is a rather subjective measurement of performance, especially in the presence of a sensorimotor impairment. Here, we propose a deep convolutional neural network (CNN) that uses electroencephalography (EEG) as an objective measurement of two kinematics components that are typically used to assess motor learning and thereby adaptation: i) the intent to initiate a goal-directed movement, and ii) the reaction …
We present a framework to generate periodic trajectory references for a 3D under-actuated bipedal robot, using a linear inverted pendulum (LIP) based controller with adaptive neural regulation. We use the LIP template model to estimate the robot's center of mass (CoM) position and velocity at the end of the current step, and formulate a discrete controller that determines the next footstep location to achieve a desired walking profile. This controller is equipped on the frontal plane with a Neural-Network-based adaptive term that reduces the model mismatch between the template and physical rob…
