Vikash Kumar

Tao Chen, Megha Tippur, Siyang Wu et al. · deepmind

In-hand object reorientation is necessary for performing many dexterous manipulation tasks, such as tool use in less structured environments that remain beyond the reach of current robots. Prior works built reorientation systems assuming one or many of the following: reorienting only specific objects with simple shapes, limited range of reorientation, slow or quasistatic manipulation, simulation-only results, the need for specialized and costly sensor suites, and other constraints which make the system infeasible for real-world deployment. We present a general object reorientation controller that does not make these assumptions. It uses readings from a single commodity depth camera to dynamically reorient complex and new object shapes by any rotation in real-time, with the median reorientation time being close to seven seconds. The controller is trained using reinforcement learning in simulation and evaluated in the real world on new object shapes not used for training, including the most challenging scenario of reorienting objects held in the air by a downward-facing hand that must counteract gravity during reorientation. Our hardware platform only uses open-source components that cost less than five thousand dollars. Although we demonstrate the ability to overcome assumptions in prior work, there is ample scope for improving absolute performance. For instance, the challenging duck-shaped object not used for training was dropped in 56 percent of the trials. When it was not dropped, our controller reoriented the object within 0.4 radians (23 degrees) 75 percent of the time. Videos are available at: https://taochenshh.github.io/projects/visual-dexterity.

Suraj Nair, Aravind Rajeswaran, Vikash Kumar et al. · stanford

We study how visual representations pre-trained on diverse human video data can enable data-efficient learning of downstream robotic manipulation tasks. Concretely, we pre-train a visual representation using the Ego4D human video dataset using a combination of time-contrastive learning, video-language alignment, and an L1 penalty to encourage sparse and compact representations. The resulting representation, R3M, can be used as a frozen perception module for downstream policy learning. Across a suite of 12 simulated robot manipulation tasks, we find that R3M improves task success by over 20% compared to training from scratch and by over 10% compared to state-of-the-art visual representations like CLIP and MoCo. Furthermore, R3M enables a Franka Emika Panda arm to learn a range of manipulation tasks in a real, cluttered apartment given just 20 demonstrations. Code and pre-trained models are available at https://tinyurl.com/robotr3m.

Tony Z. Zhao, Vikash Kumar, Sergey Levine et al.

Fine manipulation tasks, such as threading cable ties or slotting a battery, are notoriously difficult for robots because they require precision, careful coordination of contact forces, and closed-loop visual feedback. Performing these tasks typically requires high-end robots, accurate sensors, or careful calibration, which can be expensive and difficult to set up. Can learning enable low-cost and imprecise hardware to perform these fine manipulation tasks? We present a low-cost system that performs end-to-end imitation learning directly from real demonstrations, collected with a custom teleoperation interface. Imitation learning, however, presents its own challenges, particularly in high-precision domains: errors in the policy can compound over time, and human demonstrations can be non-stationary. To address these challenges, we develop a simple yet novel algorithm, Action Chunking with Transformers (ACT), which learns a generative model over action sequences. ACT allows the robot to learn 6 difficult tasks in the real world, such as opening a translucent condiment cup and slotting a battery with 80-90% success, with only 10 minutes worth of demonstrations. Project website: https://tonyzhaozh.github.io/aloha/

Albert Bou, Matteo Bettini, Sebastian Dittert et al.

PyTorch has ascended as a premier machine learning framework, yet it lacks a native and comprehensive library for decision and control tasks suitable for large development teams dealing with complex real-world data and environments. To address this issue, we propose TorchRL, a generalistic control library for PyTorch that provides well-integrated, yet standalone components. We introduce a new and flexible PyTorch primitive, the TensorDict, which facilitates streamlined algorithm development across the many branches of Reinforcement Learning (RL) and control. We provide a detailed description of the building blocks and an extensive overview of the library across domains and tasks. Finally, we experimentally demonstrate its reliability and flexibility and show comparative benchmarks to demonstrate its computational efficiency. TorchRL fosters long-term support and is publicly available on GitHub for greater reproducibility and collaboration within the research community. The code is open-sourced on GitHub.

Yecheng Jason Ma, Shagun Sodhani, Dinesh Jayaraman et al.

Reward and representation learning are two long-standing challenges for learning an expanding set of robot manipulation skills from sensory observations. Given the inherent cost and scarcity of in-domain, task-specific robot data, learning from large, diverse, offline human videos has emerged as a promising path towards acquiring a generally useful visual representation for control; however, how these human videos can be used for general-purpose reward learning remains an open question. We introduce $\textbf{V}$alue-$\textbf{I}$mplicit $\textbf{P}$re-training (VIP), a self-supervised pre-trained visual representation capable of generating dense and smooth reward functions for unseen robotic tasks. VIP casts representation learning from human videos as an offline goal-conditioned reinforcement learning problem and derives a self-supervised dual goal-conditioned value-function objective that does not depend on actions, enabling pre-training on unlabeled human videos. Theoretically, VIP can be understood as a novel implicit time contrastive objective that generates a temporally smooth embedding, enabling the value function to be implicitly defined via the embedding distance, which can then be used to construct the reward for any goal-image specified downstream task. Trained on large-scale Ego4D human videos and without any fine-tuning on in-domain, task-specific data, VIP's frozen representation can provide dense visual reward for an extensive set of simulated and $\textbf{real-robot}$ tasks, enabling diverse reward-based visual control methods and significantly outperforming all prior pre-trained representations. Notably, VIP can enable simple, $\textbf{few-shot}$ offline RL on a suite of real-world robot tasks with as few as 20 trajectories.

Kelvin Xu, Zheyuan Hu, Ria Doshi et al.

Complex and contact-rich robotic manipulation tasks, particularly those that involve multi-fingered hands and underactuated object manipulation, present a significant challenge to any control method. Methods based on reinforcement learning offer an appealing choice for such settings, as they can enable robots to learn to delicately balance contact forces and dexterously reposition objects without strong modeling assumptions. However, running reinforcement learning on real-world dexterous manipulation systems often requires significant manual engineering. This negates the benefits of autonomous data collection and ease of use that reinforcement learning should in principle provide. In this paper, we describe a system for vision-based dexterous manipulation that provides a "programming-free" approach for users to define new tasks and enable robots with complex multi-fingered hands to learn to perform them through interaction. The core principle underlying our system is that, in a vision-based setting, users should be able to provide high-level intermediate supervision that circumvents challenges in teleoperation or kinesthetic teaching which allow a robot to not only learn a task efficiently but also to autonomously practice. Our system includes a framework for users to define a final task and intermediate sub-tasks with image examples, a reinforcement learning procedure that learns the task autonomously without interventions, and experimental results with a four-finger robotic hand learning multi-stage object manipulation tasks directly in the real world, without simulation, manual modeling, or reward engineering.

Homanga Bharadhwaj, Jay Vakil, Mohit Sharma et al.

The grand aim of having a single robot that can manipulate arbitrary objects in diverse settings is at odds with the paucity of robotics datasets. Acquiring and growing such datasets is strenuous due to manual efforts, operational costs, and safety challenges. A path toward such an universal agent would require a structured framework capable of wide generalization but trained within a reasonable data budget. In this paper, we develop an efficient system (RoboAgent) for training universal agents capable of multi-task manipulation skills using (a) semantic augmentations that can rapidly multiply existing datasets and (b) action representations that can extract performant policies with small yet diverse multi-modal datasets without overfitting. In addition, reliable task conditioning and an expressive policy architecture enable our agent to exhibit a diverse repertoire of skills in novel situations specified using language commands. Using merely 7500 demonstrations, we are able to train a single agent capable of 12 unique skills, and demonstrate its generalization over 38 tasks spread across common daily activities in diverse kitchen scenes. On average, RoboAgent outperforms prior methods by over 40% in unseen situations while being more sample efficient and being amenable to capability improvements and extensions through fine-tuning. Videos at https://robopen.github.io/

Allan Zhou, Vikash Kumar, Chelsea Finn et al.

Many tasks in control, robotics, and planning can be specified using desired goal configurations for various entities in the environment. Learning goal-conditioned policies is a natural paradigm to solve such tasks. However, current approaches struggle to learn and generalize as task complexity increases, such as variations in number of environment entities or compositions of goals. In this work, we introduce a framework for modeling entity-based compositional structure in tasks, and create suitable policy designs that can leverage this structure. Our policies, which utilize architectures like Deep Sets and Self Attention, are flexible and can be trained end-to-end without requiring any action primitives. When trained using standard reinforcement and imitation learning methods on a suite of simulated robot manipulation tasks, we find that these architectures achieve significantly higher success rates with less data. We also find these architectures enable broader and compositional generalization, producing policies that extrapolate to different numbers of entities than seen in training, and stitch together (i.e. compose) learned skills in novel ways. Videos of the results can be found at https://sites.google.com/view/comp-gen-rl.

Yecheng Jason Ma, William Liang, Vaidehi Som et al.

We present Language-Image Value learning (LIV), a unified objective for vision-language representation and reward learning from action-free videos with text annotations. Exploiting a novel connection between dual reinforcement learning and mutual information contrastive learning, the LIV objective trains a multi-modal representation that implicitly encodes a universal value function for tasks specified as language or image goals. We use LIV to pre-train the first control-centric vision-language representation from large human video datasets such as EpicKitchen. Given only a language or image goal, the pre-trained LIV model can assign dense rewards to each frame in videos of unseen robots or humans attempting that task in unseen environments. Further, when some target domain-specific data is available, the same objective can be used to fine-tune and improve LIV and even other pre-trained representations for robotic control and reward specification in that domain. In our experiments on several simulated and real-world robot environments, LIV models consistently outperform the best prior input state representations for imitation learning, as well as reward specification methods for policy synthesis. Our results validate the advantages of joint vision-language representation and reward learning within the unified, compact LIV framework.

Karl Pertsch, Ruta Desai, Vikash Kumar et al.

We propose an approach for semantic imitation, which uses demonstrations from a source domain, e.g. human videos, to accelerate reinforcement learning (RL) in a different target domain, e.g. a robotic manipulator in a simulated kitchen. Instead of imitating low-level actions like joint velocities, our approach imitates the sequence of demonstrated semantic skills like "opening the microwave" or "turning on the stove". This allows us to transfer demonstrations across environments (e.g. real-world to simulated kitchen) and agent embodiments (e.g. bimanual human demonstration to robotic arm). We evaluate on three challenging cross-domain learning problems and match the performance of demonstration-accelerated RL approaches that require in-domain demonstrations. In a simulated kitchen environment, our approach learns long-horizon robot manipulation tasks, using less than 3 minutes of human video demonstrations from a real-world kitchen. This enables scaling robot learning via the reuse of demonstrations, e.g. collected as human videos, for learning in any number of target domains.

Gaoyue Zhou, Liyiming Ke, Siddhartha Srinivasa et al.

Offline reinforcement learning (ORL) holds great promise for robot learning due to its ability to learn from arbitrary pre-generated experience. However, current ORL benchmarks are almost entirely in simulation and utilize contrived datasets like replay buffers of online RL agents or sub-optimal trajectories, and thus hold limited relevance for real-world robotics. In this work (Real-ORL), we posit that data collected from safe operations of closely related tasks are more practical data sources for real-world robot learning. Under these settings, we perform an extensive (6500+ trajectories collected over 800+ robot hours and 270+ human labor hour) empirical study evaluating generalization and transfer capabilities of representative ORL methods on four real-world tabletop manipulation tasks. Our study finds that ORL and imitation learning prefer different action spaces, and that ORL algorithms can generalize from leveraging offline heterogeneous data sources and outperform imitation learning. We release our dataset and implementations at URL: https://sites.google.com/view/real-orl

Vittorio Caggiano, Huawei Wang, Guillaume Durandau et al.

Embodied agents in continuous control domains have had limited exposure to tasks allowing to explore musculoskeletal properties that enable agile and nimble behaviors in biological beings. The sophistication behind neuro-musculoskeletal control can pose new challenges for the motor learning community. At the same time, agents solving complex neural control problems allow impact in fields such as neuro-rehabilitation, as well as collaborative-robotics. Human biomechanics underlies complex multi-joint-multi-actuator musculoskeletal systems. The sensory-motor system relies on a range of sensory-contact rich and proprioceptive inputs that define and condition muscle actuation required to exhibit intelligent behaviors in the physical world. Current frameworks for musculoskeletal control do not support physiological sophistication of the musculoskeletal systems along with physical world interaction capabilities. In addition, they are neither embedded in complex and skillful motor tasks nor are computationally effective and scalable to study large-scale learning paradigms. Here, we present MyoSuite -- a suite of physiologically accurate biomechanical models of elbow, wrist, and hand, with physical contact capabilities, which allow learning of complex and skillful contact-rich real-world tasks. We provide diverse motor-control challenges: from simple postural control to skilled hand-object interactions such as turning a key, twirling a pen, rotating two balls in one hand, etc. By supporting physiological alterations in musculoskeletal geometry (tendon transfer), assistive devices (exoskeleton assistance), and muscle contraction dynamics (muscle fatigue, sarcopenia), we present real-life tasks with temporal changes, thereby exposing realistic non-stationary conditions in our tasks which most continuous control benchmarks lack.

Raunaq Bhirangi, Abigail DeFranco, Jacob Adkins et al.

High cost and lack of reliability has precluded the widespread adoption of dexterous hands in robotics. Furthermore, the lack of a viable tactile sensor capable of sensing over the entire area of the hand impedes the rich, low-level feedback that would improve learning of dexterous manipulation skills. This paper introduces an inexpensive, modular, robust, and scalable platform -- the DManus -- aimed at resolving these challenges while satisfying the large-scale data collection capabilities demanded by deep robot learning paradigms. Studies on human manipulation point to the criticality of low-level tactile feedback in performing everyday dexterous tasks. The DManus comes with ReSkin sensing on the entire surface of the palm as well as the fingertips. We demonstrate effectiveness of the fully integrated system in a tactile aware task -- bin picking and sorting. Code, documentation, design files, detailed assembly instructions, trained models, task videos, and all supplementary materials required to recreate the setup can be found on https://sites.google.com/view/roboticsbenchmarks/platforms/dmanus.

Zhao Mandi, Homanga Bharadhwaj, Vincent Moens et al.

Large-scale training have propelled significant progress in various sub-fields of AI such as computer vision and natural language processing. However, building robot learning systems at a comparable scale remains challenging. To develop robots that can perform a wide range of skills and adapt to new scenarios, efficient methods for collecting vast and diverse amounts of data on physical robot systems are required, as well as the capability to train high-capacity policies using such datasets. In this work, we propose a framework for scaling robot learning, with specific focus on multi-task and multi-scene manipulation in kitchen environments, both in simulation and in the real world. Our proposed framework, CACTI, comprises four stages that separately handle: data collection, data augmentation, visual representation learning, and imitation policy training, to enable scalability in robot learning . We make use of state-of-the-art generative models as part of the data augmentation stage, and use pre-trained out-of-domain visual representations to improve training efficiency. Experimental results demonstrate the effectiveness of our approach. On a real robot setup, CACTI enables efficient training of a single policy that can perform 10 manipulation tasks involving kitchen objects, and is robust to varying layouts of distractors. In a simulated kitchen environment, CACTI trains a single policy to perform 18 semantic tasks across 100 layout variations for each individual task. We will release the simulation task benchmark and augmented datasets in both real and simulated environments to facilitate future research.

Yuchen Cui, Scott Niekum, Abhinav Gupta et al.

Task specification is at the core of programming autonomous robots. A low-effort modality for task specification is critical for engagement of non-expert end-users and ultimate adoption of personalized robot agents. A widely studied approach to task specification is through goals, using either compact state vectors or goal images from the same robot scene. The former is hard to interpret for non-experts and necessitates detailed state estimation and scene understanding. The latter requires the generation of desired goal image, which often requires a human to complete the task, defeating the purpose of having autonomous robots. In this work, we explore alternate and more general forms of goal specification that are expected to be easier for humans to specify and use such as images obtained from the internet, hand sketches that provide a visual description of the desired task, or simple language descriptions. As a preliminary step towards this, we investigate the capabilities of large scale pre-trained models (foundation models) for zero-shot goal specification, and find promising results in a collection of simulated robot manipulation tasks and real-world datasets.

Homanga Bharadhwaj, Abhinav Gupta, Shubham Tulsiani et al.

Can we learn robot manipulation for everyday tasks, only by watching videos of humans doing arbitrary tasks in different unstructured settings? Unlike widely adopted strategies of learning task-specific behaviors or direct imitation of a human video, we develop a a framework for extracting agent-agnostic action representations from human videos, and then map it to the agent's embodiment during deployment. Our framework is based on predicting plausible human hand trajectories given an initial image of a scene. After training this prediction model on a diverse set of human videos from the internet, we deploy the trained model zero-shot for physical robot manipulation tasks, after appropriate transformations to the robot's embodiment. This simple strategy lets us solve coarse manipulation tasks like opening and closing drawers, pushing, and tool use, without access to any in-domain robot manipulation trajectories. Our real-world deployment results establish a strong baseline for action prediction information that can be acquired from diverse arbitrary videos of human activities, and be useful for zero-shot robotic manipulation in unseen scenes.

Qingpeng Zhu, Wenxiu Sun, Yuekun Dai et al.

Depth completion from RGB images and sparse Time-of-Flight (ToF) measurements is an important problem in computer vision and robotics. While traditional methods for depth completion have relied on stereo vision or structured light techniques, recent advances in deep learning have enabled more accurate and efficient completion of depth maps from RGB images and sparse ToF measurements. To evaluate the performance of different depth completion methods, we organized an RGB+sparse ToF depth completion competition. The competition aimed to encourage research in this area by providing a standardized dataset and evaluation metrics to compare the accuracy of different approaches. In this report, we present the results of the competition and analyze the strengths and weaknesses of the top-performing methods. We also discuss the implications of our findings for future research in RGB+sparse ToF depth completion. We hope that this competition and report will help to advance the state-of-the-art in this important area of research. More details of this challenge and the link to the dataset can be found at https://mipi-challenge.org/MIPI2023.

Nicklas Hansen, Yixin Lin, Hao Su et al.

Poor sample efficiency continues to be the primary challenge for deployment of deep Reinforcement Learning (RL) algorithms for real-world applications, and in particular for visuo-motor control. Model-based RL has the potential to be highly sample efficient by concurrently learning a world model and using synthetic rollouts for planning and policy improvement. However, in practice, sample-efficient learning with model-based RL is bottlenecked by the exploration challenge. In this work, we find that leveraging just a handful of demonstrations can dramatically improve the sample-efficiency of model-based RL. Simply appending demonstrations to the interaction dataset, however, does not suffice. We identify key ingredients for leveraging demonstrations in model learning -- policy pretraining, targeted exploration, and oversampling of demonstration data -- which forms the three phases of our model-based RL framework. We empirically study three complex visuo-motor control domains and find that our method is 150%-250% more successful in completing sparse reward tasks compared to prior approaches in the low data regime (100K interaction steps, 5 demonstrations). Code and videos are available at: https://nicklashansen.github.io/modemrl

Sudeep Dasari, Abhinav Gupta, Vikash Kumar

Learning diverse dexterous manipulation behaviors with assorted objects remains an open grand challenge. While policy learning methods offer a powerful avenue to attack this problem, they require extensive per-task engineering and algorithmic tuning. This paper seeks to escape these constraints, by developing a Pre-Grasp informed Dexterous Manipulation (PGDM) framework that generates diverse dexterous manipulation behaviors, without any task-specific reasoning or hyper-parameter tuning. At the core of PGDM is a well known robotics construct, pre-grasps (i.e. the hand-pose preparing for object interaction). This simple primitive is enough to induce efficient exploration strategies for acquiring complex dexterous manipulation behaviors. To exhaustively verify these claims, we introduce TCDM, a benchmark of 50 diverse manipulation tasks defined over multiple objects and dexterous manipulators. Tasks for TCDM are defined automatically using exemplar object trajectories from various sources (animators, human behaviors, etc.), without any per-task engineering and/or supervision. Our experiments validate that PGDM's exploration strategy, induced by a surprisingly simple ingredient (single pre-grasp pose), matches the performance of prior methods, which require expensive per-task feature/reward engineering, expert supervision, and hyper-parameter tuning. For animated visualizations, trained policies, and project code, please refer to: https://pregrasps.github.io/

Patrick Lancaster, Nicklas Hansen, Aravind Rajeswaran et al.

Robotic systems that aspire to operate in uninstrumented real-world environments must perceive the world directly via onboard sensing. Vision-based learning systems aim to eliminate the need for environment instrumentation by building an implicit understanding of the world based on raw pixels, but navigating the contact-rich high-dimensional search space from solely sparse visual reward signals significantly exacerbates the challenge of exploration. The applicability of such systems is thus typically restricted to simulated or heavily engineered environments since agent exploration in the real-world without the guidance of explicit state estimation and dense rewards can lead to unsafe behavior and safety faults that are catastrophic. In this study, we isolate the root causes behind these limitations to develop a system, called MoDem-V2, capable of learning contact-rich manipulation directly in the uninstrumented real world. Building on the latest algorithmic advancements in model-based reinforcement learning (MBRL), demo-bootstrapping, and effective exploration, MoDem-V2 can acquire contact-rich dexterous manipulation skills directly in the real world. We identify key ingredients for leveraging demonstrations in model learning while respecting real-world safety considerations -- exploration centering, agency handover, and actor-critic ensembles. We empirically demonstrate the contribution of these ingredients in four complex visuo-motor manipulation problems in both simulation and the real world. To the best of our knowledge, our work presents the first successful system for demonstration-augmented visual MBRL trained directly in the real world. Visit https://sites.google.com/view/modem-v2 for videos and more details.

Cameron Berg, Vittorio Caggiano, Vikash Kumar

Learning effective continuous control policies in high-dimensional systems, including musculoskeletal agents, remains a significant challenge. Over the course of biological evolution, organisms have developed robust mechanisms for overcoming this complexity to learn highly sophisticated strategies for motor control. What accounts for this robust behavioral flexibility? Modular control via muscle synergies, i.e. coordinated muscle co-contractions, is considered to be one putative mechanism that enables organisms to learn muscle control in a simplified and generalizable action space. Drawing inspiration from this evolved motor control strategy, we use physiologically accurate human hand and leg models as a testbed for determining the extent to which a Synergistic Action Representation (SAR) acquired from simpler tasks facilitates learning more complex tasks. We find in both cases that SAR-exploiting policies significantly outperform end-to-end reinforcement learning. Policies trained with SAR were able to achieve robust locomotion on a wide set of terrains with high sample efficiency, while baseline approaches failed to learn meaningful behaviors. Additionally, policies trained with SAR on a multiobject manipulation task significantly outperformed (>70% success) baseline approaches (<20% success). Both of these SAR-exploiting policies were also found to generalize zero-shot to out-of-domain environmental conditions, while policies that did not adopt SAR failed to generalize. Finally, we establish the generality of SAR on broader high-dimensional control problems using a robotic manipulation task set and a full-body humanoid locomotion task. To the best of our knowledge, this investigation is the first of its kind to present an end-to-end pipeline for discovering synergies and using this representation to learn high-dimensional continuous control across a wide diversity of tasks.

Vittorio Caggiano, Sudeep Dasari, Vikash Kumar

Human dexterity is a hallmark of motor control. Our hands can rapidly synthesize new behaviors despite the complexity (multi-articular and multi-joints, with 23 joints controlled by more than 40 muscles) of musculoskeletal sensory-motor circuits. In this work, we take inspiration from how human dexterity builds on a diversity of prior experiences, instead of being acquired through a single task. Motivated by this observation, we set out to develop agents that can build upon their previous experience to quickly acquire new (previously unattainable) behaviors. Specifically, our approach leverages multi-task learning to implicitly capture task-agnostic behavioral priors (MyoDex) for human-like dexterity, using a physiologically realistic human hand model - MyoHand. We demonstrate MyoDex's effectiveness in few-shot generalization as well as positive transfer to a large repertoire of unseen dexterous manipulation tasks. Agents leveraging MyoDex can solve approximately 3x more tasks, and 4x faster in comparison to a distillation baseline. While prior work has synthesized single musculoskeletal control behaviors, MyoDex is the first generalizable manipulation prior that catalyzes the learning of dexterous physiological control across a large variety of contact-rich behaviors. We also demonstrate the effectiveness of our paradigms beyond musculoskeletal control towards the acquisition of dexterity in 24 DoF Adroit Hand. Website: https://sites.google.com/view/myodex

Zheyuan Hu, Aaron Rovinsky, Jianlan Luo et al.

Dexterous manipulation tasks involving contact-rich interactions pose a significant challenge for both model-based control systems and imitation learning algorithms. The complexity arises from the need for multi-fingered robotic hands to dynamically establish and break contacts, balance non-prehensile forces, and control large degrees of freedom. Reinforcement learning (RL) offers a promising approach due to its general applicability and capacity to autonomously acquire optimal manipulation strategies. However, its real-world application is often hindered by the necessity to generate a large number of samples, reset the environment, and obtain reward signals. In this work, we introduce an efficient system for learning dexterous manipulation skills with RL to alleviate these challenges. The main idea of our approach is the integration of recent advances in sample-efficient RL and replay buffer bootstrapping. This combination allows us to utilize data from different tasks or objects as a starting point for training new tasks, significantly improving learning efficiency. Additionally, our system completes the real-world training cycle by incorporating learned resets via an imitation-based pickup policy as well as learned reward functions, eliminating the need for manual resets and reward engineering. We demonstrate the benefits of reusing past data as replay buffer initialization for new tasks, for instance, the fast acquisition of intricate manipulation skills in the real world on a four-fingered robotic hand. (Videos: https://sites.google.com/view/reboot-dexterous)

Zoey Chen, Zhao Mandi, Homanga Bharadhwaj et al.

Generalization to unseen real-world scenarios for robot manipulation requires exposure to diverse datasets during training. However, collecting large real-world datasets is intractable due to high operational costs. For robot learning to generalize despite these challenges, it is essential to leverage sources of data or priors beyond the robot's direct experience. In this work, we posit that image-text generative models, which are pre-trained on large corpora of web-scraped data, can serve as such a data source. These generative models encompass a broad range of real-world scenarios beyond a robot's direct experience and can synthesize novel synthetic experiences that expose robotic agents to additional world priors aiding real-world generalization at no extra cost. In particular, our approach leverages pre-trained generative models as an effective tool for data augmentation. We propose a generative augmentation framework for semantically controllable augmentations and rapidly multiplying robot datasets while inducing rich variations that enable real-world generalization. Based on diverse augmentations of robot data, we show how scalable robot manipulation policies can be trained and deployed both in simulation and in unseen real-world environments such as kitchens and table-tops. By demonstrating the effectiveness of image-text generative models in diverse real-world robotic applications, our generative augmentation framework provides a scalable and efficient path for boosting generalization in robot learning at no extra human cost.

Cheryl Wang, Chun Kwang Tan, Balint K. Hodossy et al.

Athletic performance represents the pinnacle of human motor intelligence, demanding rapid choices, precise control, agility, and coordinated physical execution. Replicating this seamless combination of capabilities remains elusive in current artificial intelligence and robotic systems. Concurrently, understanding the biological mastery of these movements is hindered because complex muscle coordination is rarely measured in vivo due to the limitations of physical equipment. To bridge this fundamental gap in understanding, MyoChallenge at NeurIPS 2025 established a pioneering benchmark for motor control intelligence in sports, leveraging high-fidelity musculoskeletal models within physics simulation combined with machine learning-driven algorithms. The competition introduces two distinct tracks emphasizing either upper or lower limbs control: a table tennis rally task utilizing a biomechanic upper limb composed of an arm with a hand and a trunk; and a soccer penalty kick using a biomechanic model of legs and a trunk. Marking the fourth iteration of the MyoChallenge series, this event attracted almost 70 teams and over 560 submissions globally, uniting a diverse community ranging from physicians and neuroscientists to machine learning experts. The competition facilitated the development of several state-of-the-art control algorithms for a musculoskeletal system capable of sports agility, leveraging techniques such as physics-based motion planners, on-policy behaviour cloning, hierarchical planning, and muscle synergies. By integrating standardized tasks and physiologically realistic models into the open-source framework of MyoSuite, MyoChallenge'25 serves as a reproducible and reusable testbed to accelerate interdisciplinary research across machine learning, biomechanics, sports science, and neuroscience. Project page: https://www.myosuite.org//myochallenge/myochallenge-2025.

Pierre Schumacher, Thomas Geijtenbeek, Vittorio Caggiano et al.

Humans excel at robust bipedal walking in complex natural environments. In each step, they adequately tune the interaction of biomechanical muscle dynamics and neuronal signals to be robust against uncertainties in ground conditions. However, it is still not fully understood how the nervous system resolves the musculoskeletal redundancy to solve the multi-objective control problem considering stability, robustness, and energy efficiency. In computer simulations, energy minimization has been shown to be a successful optimization target, reproducing natural walking with trajectory optimization or reflex-based control methods. However, these methods focus on particular motions at a time and the resulting controllers are limited when compensating for perturbations. In robotics, reinforcement learning~(RL) methods recently achieved highly stable (and efficient) locomotion on quadruped systems, but the generation of human-like walking with bipedal biomechanical models has required extensive use of expert data sets. This strong reliance on demonstrations often results in brittle policies and limits the application to new behaviors, especially considering the potential variety of movements for high-dimensional musculoskeletal models in 3D. Achieving natural locomotion with RL without sacrificing its incredible robustness might pave the way for a novel approach to studying human walking in complex natural environments. Videos: https://sites.google.com/view/naturalwalkingrl

Vikash Kumar, Rohit Lal, Himanshu Patil et al.

This work introduces the novel task of Source-free Multi-target Domain Adaptation and proposes adaptation framework comprising of \textbf{Co}nsistency with \textbf{N}uclear-Norm Maximization and \textbf{Mix}Up knowledge distillation (\textit{CoNMix}) as a solution to this problem. The main motive of this work is to solve for Single and Multi target Domain Adaptation (SMTDA) for the source-free paradigm, which enforces a constraint where the labeled source data is not available during target adaptation due to various privacy-related restrictions on data sharing. The source-free approach leverages target pseudo labels, which can be noisy, to improve the target adaptation. We introduce consistency between label preserving augmentations and utilize pseudo label refinement methods to reduce noisy pseudo labels. Further, we propose novel MixUp Knowledge Distillation (MKD) for better generalization on multiple target domains using various source-free STDA models. We also show that the Vision Transformer (VT) backbone gives better feature representation with improved domain transferability and class discriminability. Our proposed framework achieves the state-of-the-art (SOTA) results in various paradigms of source-free STDA and MTDA settings on popular domain adaptation datasets like Office-Home, Office-Caltech, and DomainNet. Project Page: https://sites.google.com/view/conmix-vcl

Maijunxian Wang, Ruisi Wang, Juyi Lin et al.

Rapid progress in video models has largely focused on visual quality, leaving their reasoning capabilities underexplored. Video reasoning grounds intelligence in spatiotemporally consistent visual environments that go beyond what text can naturally capture, enabling intuitive reasoning over spatiotemporal structure such as continuity, interaction, and causality. However, systematically studying video reasoning and its scaling behavior is hindered by the lack of large-scale training data. To address this gap, we introduce the Very Big Video Reasoning (VBVR) Dataset, an unprecedentedly large-scale resource spanning 200 curated reasoning tasks following a principled taxonomy and over one million video clips, approximately three orders of magnitude larger than existing datasets. We further present VBVR-Bench, a verifiable evaluation framework that moves beyond model-based judging by incorporating rule-based, human-aligned scorers, enabling reproducible and interpretable diagnosis of video reasoning capabilities. Leveraging the VBVR suite, we conduct one of the first large-scale scaling studies of video reasoning and observe early signs of emergent generalization to unseen reasoning tasks. Together, VBVR lays a foundation for the next stage of research in generalizable video reasoning. The data, benchmark toolkit, and models are publicly available at https://video-reason.com/ .

Vikash Kumar, Himanshu Patil, Rohit Lal et al.

In this work, we explore the usage of the Frequency Transformation for reducing the domain shift between the source and target domain (e.g., synthetic image and real image respectively) towards solving the Domain Adaptation task. Most of the Unsupervised Domain Adaptation (UDA) algorithms focus on reducing the global domain shift between labelled source and unlabelled target domains by matching the marginal distributions under a small domain gap assumption. UDA performance degrades for the cases where the domain gap between source and target distribution is large. In order to bring the source and the target domains closer, we propose a novel approach based on traditional image processing technique Class Aware Frequency Transformation (CAFT) that utilizes pseudo label based class consistent low-frequency swapping for improving the overall performance of the existing UDA algorithms. The proposed approach, when compared with the state-of-the-art deep learning based methods, is computationally more efficient and can easily be plugged into any existing UDA algorithm to improve its performance. Additionally, we introduce a novel approach based on absolute difference of top-2 class prediction probabilities (ADT2P) for filtering target pseudo labels into clean and noisy sets. Samples with clean pseudo labels can be used to improve the performance of unsupervised learning algorithms. We name the overall framework as CAFT++. We evaluate the same on the top of different UDA algorithms across many public domain adaptation datasets. Our extensive experiments indicate that CAFT++ is able to achieve significant performance gains across all the popular benchmarks.

Vikash Kumar, Subrata Mukherjee, Somnath Sarangi

Gearbox fault diagnosis is one of the most important parts in any industrial systems. Failure of components inside gearbox can lead to a catastrophic failure, uneven breakdown, and financial losses in industrial organization. In that case intelligent maintenance of the gearbox comes into context. This paper presents an integrated gearbox fault diagnosis approach which can easily deploy in online condition monitoring. This work introduces a nonparametric data preprocessing technique i.e., calculus enhanced energy operator (CEEO) to preserve the characteristics frequencies in the noisy and inferred vibrational signal. A set of time domain and spectral domain features are calculated from the raw and CEEO vibration signal and inputted to the multiclass support vector machine (MCSVM) to diagnose the faults on the system. An effective comparison between raw signal and CEEO signal are presented to show the impact of CEEO in gearbox fault diagnosis. The obtained results of this work look very promising and can be implemented in any type of industrial system due to its nonparametric nature.

Subrata Mukherjee, Vikash Kumar, Somnath Sarangi

This paper discusses the possible observation of an integrated gear tooth crack analysis procedure that employs the combined approach of variable mode decomposition (VMD) and time synchronous averaging (TSA) based on the coupled electromechanical gearbox (CEMG) system. This paper also incorporates the modified Lagrangian formulation to model the CEMG system by considering Rayleigh's dissipative potential. An analytical improved time-varying mesh stiffness (IAM-TVMS) with different levels of gear tooth crack depts is also incorporated into the CEMG system to inspect the influence of cracks on the system's dynamic behavior. Dynamic responses of the CEMG system with different tooth crack levels have been used for further investigations. For the first time, the integrated approach of variable mode decomposition (VMD) and time-synchronous averaging (TSA) has been presented to analyze the dynamic behaviour of CEMG systems at the different gear tooth cracks have been experienced as non-stationary and complex vibration signals with noise. Based on the integrated approach of VMD-TSA, two types of nonlinear features, i.e., Lyapunov Exponent (LE) and Correlation Dimension (CD), were calculated to predict the level of chaotic vibration and complexity of the CEMG system at the different levels of gear tooth cracks. Also, the LE and CD are used as chaotic behaviour features to predict the gear tooth crack propagation level. The results of the proposed approach show significant improvements in the gear tooth crack analysis based on the chaotic features. Also, this is one of the first attempts to study the CEMG system using chaotic features based on the combined approach of VMD-TSA.

Michael Ahn, Henry Zhu, Kristian Hartikainen et al.

ROBEL is an open-source platform of cost-effective robots designed for reinforcement learning in the real world. ROBEL introduces two robots, each aimed to accelerate reinforcement learning research in different task domains: D'Claw is a three-fingered hand robot that facilitates learning dexterous manipulation tasks, and D'Kitty is a four-legged robot that facilitates learning agile legged locomotion tasks. These low-cost, modular robots are easy to maintain and are robust enough to sustain on-hardware reinforcement learning from scratch with over 14000 training hours registered on them to date. To leverage this platform, we propose an extensible set of continuous control benchmark tasks for each robot. These tasks feature dense and sparse task objectives, and additionally introduce score metrics as hardware-safety. We provide benchmark scores on an initial set of tasks using a variety of learning-based methods. Furthermore, we show that these results can be replicated across copies of the robots located in different institutions. Code, documentation, design files, detailed assembly instructions, final policies, baseline details, task videos, and all supplementary materials required to reproduce the results are available at www.roboticsbenchmarks.org.

Mohit Sharma, Adam Fishman, Vikash Kumar et al.

Robots in the real world need to perceive and move to goals in complex environments without collisions. Avoiding collisions is especially difficult when relying on sensor perception and when goals are among clutter. Diffusion policies and other generative models have shown strong performance in solving local planning problems, but often struggle at avoiding all of the subtle constraint violations that characterize truly challenging global motion planning problems. In this work, we propose an approach for learning global motion planning using diffusion policies, allowing the robot to generate full trajectories through complex scenes and reasoning about multiple obstacles along the path. Our approach uses cascaded hierarchical models which unify global prediction and local refinement together with online plan repair to ensure the trajectories are collision free. Our method outperforms (by ~5%) a wide variety of baselines on challenging tasks in multiple domains including navigation and manipulation.

Vikash Kumar

Estimating Worst-Case Execution Time (WCET) is of utmost importance for developing Cyber-Physical and Safety-Critical Systems. The system's scheduler uses the estimated WCET to schedule each task of these systems, and failure may lead to catastrophic events. It is thus imperative to build provably reliable systems. WCET is available to us in the last stage of systems development when the hardware is available and the application code is compiled on it. Different methodologies measure the WCET, but none of them give early insights on WCET, which is crucial for system development. If the system designers overestimate WCET in the early stage, then it would lead to the overqualified system, which will increase the cost of the final product, and if they underestimate WCET in the early stage, then it would lead to financial loss as the system would not perform as expected. This paper estimates early WCET using Deep Neural Networks as an approximate predictor model for hardware architecture and compiler. This model predicts the WCET based on the source code without compiling and running on the hardware architecture. Our WCET prediction model is created using the Pytorch framework. The resulting WCET is too erroneous to be used as an upper bound on the WCET. However, getting these results in the early stages of system development is an essential prerequisite for the system's dimensioning and configuration of the hardware setup.

Rutav Shah, Vikash Kumar

The ability to autonomously learn behaviors via direct interactions in uninstrumented environments can lead to generalist robots capable of enhancing productivity or providing care in unstructured settings like homes. Such uninstrumented settings warrant operations only using the robot's proprioceptive sensor such as onboard cameras, joint encoders, etc which can be challenging for policy learning owing to the high dimensionality and partial observability issues. We propose RRL: Resnet as representation for Reinforcement Learning -- a straightforward yet effective approach that can learn complex behaviors directly from proprioceptive inputs. RRL fuses features extracted from pre-trained Resnet into the standard reinforcement learning pipeline and delivers results comparable to learning directly from the state. In a simulated dexterous manipulation benchmark, where the state of the art methods fail to make significant progress, RRL delivers contact rich behaviors. The appeal of RRL lies in its simplicity in bringing together progress from the fields of Representation Learning, Imitation Learning, and Reinforcement Learning. Its effectiveness in learning behaviors directly from visual inputs with performance and sample efficiency matching learning directly from the state, even in complex high dimensional domains, is far from obvious.

Abhishek Gupta, Justin Yu, Tony Z. Zhao et al.

Reinforcement Learning (RL) algorithms can in principle acquire complex robotic skills by learning from large amounts of data in the real world, collected via trial and error. However, most RL algorithms use a carefully engineered setup in order to collect data, requiring human supervision and intervention to provide episodic resets. This is particularly evident in challenging robotics problems, such as dexterous manipulation. To make data collection scalable, such applications require reset-free algorithms that are able to learn autonomously, without explicit instrumentation or human intervention. Most prior work in this area handles single-task learning. However, we might also want robots that can perform large repertoires of skills. At first, this would appear to only make the problem harder. However, the key observation we make in this work is that an appropriately chosen multi-task RL setting actually alleviates the reset-free learning challenge, with minimal additional machinery required. In effect, solving a multi-task problem can directly solve the reset-free problem since different combinations of tasks can serve to perform resets for other tasks. By learning multiple tasks together and appropriately sequencing them, we can effectively learn all of the tasks together reset-free. This type of multi-task learning can effectively scale reset-free learning schemes to much more complex problems, as we demonstrate in our experiments. We propose a simple scheme for multi-task learning that tackles the reset-free learning problem, and show its effectiveness at learning to solve complex dexterous manipulation tasks in both hardware and simulation without any explicit resets. This work shows the ability to learn dexterous manipulation behaviors in the real world with RL without any human intervention.

Archit Sharma, Michael Ahn, Sergey Levine et al.

Reinforcement learning provides a general framework for learning robotic skills while minimizing engineering effort. However, most reinforcement learning algorithms assume that a well-designed reward function is provided, and learn a single behavior for that single reward function. Such reward functions can be difficult to design in practice. Can we instead develop efficient reinforcement learning methods that acquire diverse skills without any reward function, and then repurpose these skills for downstream tasks? In this paper, we demonstrate that a recently proposed unsupervised skill discovery algorithm can be extended into an efficient off-policy method, making it suitable for performing unsupervised reinforcement learning in the real world. Firstly, we show that our proposed algorithm provides substantial improvement in learning efficiency, making reward-free real-world training feasible. Secondly, we move beyond the simulation environments and evaluate the algorithm on real physical hardware. On quadrupeds, we observe that locomotion skills with diverse gaits and different orientations emerge without any rewards or demonstrations. We also demonstrate that the learned skills can be composed using model predictive control for goal-oriented navigation, without any additional training.

Henry Zhu, Justin Yu, Abhishek Gupta et al.

The success of reinforcement learning for real world robotics has been, in many cases limited to instrumented laboratory scenarios, often requiring arduous human effort and oversight to enable continuous learning. In this work, we discuss the elements that are needed for a robotic learning system that can continually and autonomously improve with data collected in the real world. We propose a particular instantiation of such a system, using dexterous manipulation as our case study. Subsequently, we investigate a number of challenges that come up when learning without instrumentation. In such settings, learning must be feasible without manually designed resets, using only on-board perception, and without hand-engineered reward functions. We propose simple and scalable solutions to these challenges, and then demonstrate the efficacy of our proposed system on a set of dexterous robotic manipulation tasks, providing an in-depth analysis of the challenges associated with this learning paradigm. We demonstrate that our complete system can learn without any human intervention, acquiring a variety of vision-based skills with a real-world three-fingered hand. Results and videos can be found at https://sites.google.com/view/realworld-rl/

Aravind Rajeswaran, Igor Mordatch, Vikash Kumar

Model-based reinforcement learning (MBRL) has recently gained immense interest due to its potential for sample efficiency and ability to incorporate off-policy data. However, designing stable and efficient MBRL algorithms using rich function approximators have remained challenging. To help expose the practical challenges in MBRL and simplify algorithm design from the lens of abstraction, we develop a new framework that casts MBRL as a game between: (1) a policy player, which attempts to maximize rewards under the learned model; (2) a model player, which attempts to fit the real-world data collected by the policy player. For algorithm development, we construct a Stackelberg game between the two players, and show that it can be solved with approximate bi-level optimization. This gives rise to two natural families of algorithms for MBRL based on which player is chosen as the leader in the Stackelberg game. Together, they encapsulate, unify, and generalize many previous MBRL algorithms. Furthermore, our framework is consistent with and provides a clear basis for heuristics known to be important in practice from prior works. Finally, through experiments we validate that our proposed algorithms are highly sample efficient, match the asymptotic performance of model-free policy gradient, and scale gracefully to high-dimensional tasks like dexterous hand manipulation. Additional details and code can be obtained from the project page at https://sites.google.com/view/mbrl-game

Silvia Cruciani, Balakumar Sundaralingam, Kaiyu Hang et al.

The purpose of this benchmark is to evaluate the planning and control aspects of robotic in-hand manipulation systems. The goal is to assess the system's ability to change the pose of a hand-held object by either using the fingers, environment or a combination of both. Given an object surface mesh from the YCB data-set, we provide examples of initial and goal states (i.e.\ static object poses and fingertip locations) for various in-hand manipulation tasks. We further propose metrics that measure the error in reaching the goal state from a specific initial state, which, when aggregated across all tasks, also serves as a measure of the system's in-hand manipulation capability. We provide supporting software, task examples, and evaluation results associated with the benchmark. All the supporting material is available at https://robot-learning.cs.utah.edu/project/benchmarking_in_hand_manipulation

Abhishek Gupta, Vikash Kumar, Corey Lynch et al.

We present relay policy learning, a method for imitation and reinforcement learning that can solve multi-stage, long-horizon robotic tasks. This general and universally-applicable, two-phase approach consists of an imitation learning stage that produces goal-conditioned hierarchical policies, and a reinforcement learning phase that finetunes these policies for task performance. Our method, while not necessarily perfect at imitation learning, is very amenable to further improvement via environment interaction, allowing it to scale to challenging long-horizon tasks. We simplify the long-horizon policy learning problem by using a novel data-relabeling algorithm for learning goal-conditioned hierarchical policies, where the low-level only acts for a fixed number of steps, regardless of the goal achieved. While we rely on demonstration data to bootstrap policy learning, we do not assume access to demonstrations of every specific tasks that is being solved, and instead leverage unstructured and unsegmented demonstrations of semantically meaningful behaviors that are not only less burdensome to provide, but also can greatly facilitate further improvement using reinforcement learning. We demonstrate the effectiveness of our method on a number of multi-stage, long-horizon manipulation tasks in a challenging kitchen simulation environment. Videos are available at https://relay-policy-learning.github.io/

Anusha Nagabandi, Kurt Konoglie, Sergey Levine et al.

Dexterous multi-fingered hands can provide robots with the ability to flexibly perform a wide range of manipulation skills. However, many of the more complex behaviors are also notoriously difficult to control: Performing in-hand object manipulation, executing finger gaits to move objects, and exhibiting precise fine motor skills such as writing, all require finely balancing contact forces, breaking and reestablishing contacts repeatedly, and maintaining control of unactuated objects. Learning-based techniques provide the appealing possibility of acquiring these skills directly from data, but current learning approaches either require large amounts of data and produce task-specific policies, or they have not yet been shown to scale up to more complex and realistic tasks requiring fine motor skills. In this work, we demonstrate that our method of online planning with deep dynamics models (PDDM) addresses both of these limitations; we show that improvements in learned dynamics models, together with improvements in online model-predictive control, can indeed enable efficient and effective learning of flexible contact-rich dexterous manipulation skills -- and that too, on a 24-DoF anthropomorphic hand in the real world, using just 4 hours of purely real-world data to learn to simultaneously coordinate multiple free-floating objects. Videos can be found at https://sites.google.com/view/pddm/

Ofir Nachum, Michael Ahn, Hugo Ponte et al.

Manipulation and locomotion are closely related problems that are often studied in isolation. In this work, we study the problem of coordinating multiple mobile agents to exhibit manipulation behaviors using a reinforcement learning (RL) approach. Our method hinges on the use of hierarchical sim2real -- a simulated environment is used to learn low-level goal-reaching skills, which are then used as the action space for a high-level RL controller, also trained in simulation. The full hierarchical policy is then transferred to the real world in a zero-shot fashion. The application of domain randomization during training enables the learned behaviors to generalize to real-world settings, while the use of hierarchy provides a modular paradigm for learning and transferring increasingly complex behaviors. We evaluate our method on a number of real-world tasks, including coordinated object manipulation in a multi-agent setting. See videos at https://sites.google.com/view/manipulation-via-locomotion

Archit Sharma, Shixiang Gu, Sergey Levine et al.

Conventionally, model-based reinforcement learning (MBRL) aims to learn a global model for the dynamics of the environment. A good model can potentially enable planning algorithms to generate a large variety of behaviors and solve diverse tasks. However, learning an accurate model for complex dynamical systems is difficult, and even then, the model might not generalize well outside the distribution of states on which it was trained. In this work, we combine model-based learning with model-free learning of primitives that make model-based planning easy. To that end, we aim to answer the question: how can we discover skills whose outcomes are easy to predict? We propose an unsupervised learning algorithm, Dynamics-Aware Discovery of Skills (DADS), which simultaneously discovers predictable behaviors and learns their dynamics. Our method can leverage continuous skill spaces, theoretically, allowing us to learn infinitely many behaviors even for high-dimensional state-spaces. We demonstrate that zero-shot planning in the learned latent space significantly outperforms standard MBRL and model-free goal-conditioned RL, can handle sparse-reward tasks, and substantially improves over prior hierarchical RL methods for unsupervised skill discovery.

Corey Lynch, Mohi Khansari, Ted Xiao et al.

Acquiring a diverse repertoire of general-purpose skills remains an open challenge for robotics. In this work, we propose self-supervising control on top of human teleoperated play data as a way to scale up skill learning. Play has two properties that make it attractive compared to conventional task demonstrations. Play is cheap, as it can be collected in large quantities quickly without task segmenting, labeling, or resetting to an initial state. Play is naturally rich, covering ~4x more interaction space than task demonstrations for the same amount of collection time. To learn control from play, we introduce Play-LMP, a self-supervised method that learns to organize play behaviors in a latent space, then reuse them at test time to achieve specific goals. Combining self-supervised control with a diverse play dataset shifts the focus of skill learning from a narrow and discrete set of tasks to the full continuum of behaviors available in an environment. We find that this combination generalizes well empirically---after self-supervising on unlabeled play, our method substantially outperforms individual expert-trained policies on 18 difficult user-specified visual manipulation tasks in a simulated robotic tabletop environment. We additionally find that play-supervised models, unlike their expert-trained counterparts, are more robust to perturbations and exhibit retrying-till-success behaviors. Finally, we find that our agent organizes its latent plan space around functional tasks, despite never being trained with task labels. Videos, code and data are available at learning-from-play.github.io

Tuomas Haarnoja, Aurick Zhou, Kristian Hartikainen et al.

Model-free deep reinforcement learning (RL) algorithms have been successfully applied to a range of challenging sequential decision making and control tasks. However, these methods typically suffer from two major challenges: high sample complexity and brittleness to hyperparameters. Both of these challenges limit the applicability of such methods to real-world domains. In this paper, we describe Soft Actor-Critic (SAC), our recently introduced off-policy actor-critic algorithm based on the maximum entropy RL framework. In this framework, the actor aims to simultaneously maximize expected return and entropy. That is, to succeed at the task while acting as randomly as possible. We extend SAC to incorporate a number of modifications that accelerate training and improve stability with respect to the hyperparameters, including a constrained formulation that automatically tunes the temperature hyperparameter. We systematically evaluate SAC on a range of benchmark tasks, as well as real-world challenging tasks such as locomotion for a quadrupedal robot and robotic manipulation with a dexterous hand. With these improvements, SAC achieves state-of-the-art performance, outperforming prior on-policy and off-policy methods in sample-efficiency and asymptotic performance. Furthermore, we demonstrate that, in contrast to other off-policy algorithms, our approach is very stable, achieving similar performance across different random seeds. These results suggest that SAC is a promising candidate for learning in real-world robotics tasks.

Henry Zhu, Abhishek Gupta, Aravind Rajeswaran et al.

Dexterous multi-fingered robotic hands can perform a wide range of manipulation skills, making them an appealing component for general-purpose robotic manipulators. However, such hands pose a major challenge for autonomous control, due to the high dimensionality of their configuration space and complex intermittent contact interactions. In this work, we propose deep reinforcement learning (deep RL) as a scalable solution for learning complex, contact rich behaviors with multi-fingered hands. Deep RL provides an end-to-end approach to directly map sensor readings to actions, without the need for task specific models or policy classes. We show that contact-rich manipulation behavior with multi-fingered hands can be learned by directly training with model-free deep RL algorithms in the real world, with minimal additional assumption and without the aid of simulation. We learn a variety of complex behaviors on two different low-cost hardware platforms. We show that each task can be learned entirely from scratch, and further study how the learning process can be further accelerated by using a small number of human demonstrations to bootstrap learning. Our experiments demonstrate that complex multi-fingered manipulation skills can be learned in the real world in about 4-7 hours for most tasks, and that demonstrations can decrease this to 2-3 hours, indicating that direct deep RL training in the real world is a viable and practical alternative to simulation and model-based control. \url{https://sites.google.com/view/deeprl-handmanipulation}

Suraj Nair, Mohammad Babaeizadeh, Chelsea Finn et al.

A longstanding challenge in robot learning for manipulation tasks has been the ability to generalize to varying initial conditions, diverse objects, and changing objectives. Learning based approaches have shown promise in producing robust policies, but require heavy supervision to efficiently learn precise control, especially from visual inputs. We propose a novel self-supervision technique that uses time-reversal to learn goals and provide a high level plan to reach them. In particular, we introduce the time-reversal model (TRM), a self-supervised model which explores outward from a set of goal states and learns to predict these trajectories in reverse. This provides a high level plan towards goals, allowing us to learn complex manipulation tasks with no demonstrations or exploration at test time. We test our method on the domain of assembly, specifically the mating of tetris-style block pairs. Using our method operating atop visual model predictive control, we are able to assemble tetris blocks on a physical robot using only uncalibrated RGB camera input, and generalize to unseen block pairs. sites.google.com/view/time-reversal

Cathy Wu, Aravind Rajeswaran, Yan Duan et al.

Policy gradient methods have enjoyed great success in deep reinforcement learning but suffer from high variance of gradient estimates. The high variance problem is particularly exasperated in problems with long horizons or high-dimensional action spaces. To mitigate this issue, we derive a bias-free action-dependent baseline for variance reduction which fully exploits the structural form of the stochastic policy itself and does not make any additional assumptions about the MDP. We demonstrate and quantify the benefit of the action-dependent baseline through both theoretical analysis as well as numerical results, including an analysis of the suboptimality of the optimal state-dependent baseline. The result is a computationally efficient policy gradient algorithm, which scales to high-dimensional control problems, as demonstrated by a synthetic 2000-dimensional target matching task. Our experimental results indicate that action-dependent baselines allow for faster learning on standard reinforcement learning benchmarks and high-dimensional hand manipulation and synthetic tasks. Finally, we show that the general idea of including additional information in baselines for improved variance reduction can be extended to partially observed and multi-agent tasks.

Matthias Plappert, Marcin Andrychowicz, Alex Ray et al.

The purpose of this technical report is two-fold. First of all, it introduces a suite of challenging continuous control tasks (integrated with OpenAI Gym) based on currently existing robotics hardware. The tasks include pushing, sliding and pick & place with a Fetch robotic arm as well as in-hand object manipulation with a Shadow Dexterous Hand. All tasks have sparse binary rewards and follow a Multi-Goal Reinforcement Learning (RL) framework in which an agent is told what to do using an additional input. The second part of the paper presents a set of concrete research ideas for improving RL algorithms, most of which are related to Multi-Goal RL and Hindsight Experience Replay.