Search

Significant progress has been made in open-vocabulary mobile manipulation, where the goal is for a robot to perform tasks in any environment given a natural language description. However, most current systems assume a static environment, which limits the system's applicability in real-world scenarios where environments frequently change due to human intervention or the robot's own actions. In this work, we present DynaMem, a new approach to open-world mobile manipulation that uses a dynamic spatio-semantic memory to represent a robot's environment. DynaMem constructs a 3D data structure to maintain a dynamic memory of point clouds, and answers open-vocabulary object localization queries using multimodal LLMs or open-vocabulary features generated by state-of-the-art vision-language models. Powered by DynaMem, our robots can explore novel environments, search for objects not found in memory, and continuously update the memory as objects move, appear, or disappear in the scene. We run extensive experiments on the Stretch SE3 robots in three real and nine offline scenes, and achieve an average pick-and-drop success rate of 70% on non-stationary objects, which is more than a 2x improvement over state-of-the-art static systems. Our code as well as our experiment and deployment videos are open sourced and can be found on our project website: https://dynamem.github.io/, Comment: Website: https://dynamem.github.io

Robot models, particularly those trained with large amounts of data, have recently shown a plethora of real-world manipulation and navigation capabilities. Several independent efforts have shown that given sufficient training data in an environment, robot policies can generalize to demonstrated variations in that environment. However, needing to finetune robot models to every new environment stands in stark contrast to models in language or vision that can be deployed zero-shot for open-world problems. In this work, we present Robot Utility Models (RUMs), a framework for training and deploying zero-shot robot policies that can directly generalize to new environments without any finetuning. To create RUMs efficiently, we develop new tools to quickly collect data for mobile manipulation tasks, integrate such data into a policy with multi-modal imitation learning, and deploy policies on-device on Hello Robot Stretch, a cheap commodity robot, with an external mLLM verifier for retrying. We train five such utility models for opening cabinet doors, opening drawers, picking up napkins, picking up paper bags, and reorienting fallen objects. Our system, on average, achieves 90% success rate in unseen, novel environments interacting with unseen objects. Moreover, the utility models can also succeed in different robot and camera set-ups with no further data, training, or fine-tuning. Primary among our lessons are the importance of training data over training algorithm and policy class, guidance about data scaling, necessity for diverse yet high-quality demonstrations, and a recipe for robot introspection and retrying to improve performance on individual environments. Our code, data, models, hardware designs, as well as our experiment and deployment videos are open sourced and can be found on our project website: https://robotutilitymodels.com, Comment: Project website https://robotutilitymodels.com

De gustibus non est disputandum ("there is no accounting for others' tastes") is a common Latin maxim describing how many solutions in life are determined by people's personal preferences. Many household tasks, in particular, can only be considered fully successful when they account for personal preferences such as the visual aesthetic of the scene. For example, setting a table could be optimized by arranging utensils according to traditional rules of Western table setting decorum, without considering the color, shape, or material of each object, but this may not be a completely satisfying solution for a given person. Toward this end, we present DegustaBot, an algorithm for visual preference learning that solves household multi-object rearrangement tasks according to personal preference. To do this, we use internet-scale pre-trained vision-and-language foundation models (VLMs) with novel zero-shot visual prompting techniques. To evaluate our method, we collect a large dataset of naturalistic personal preferences in a simulated table-setting task, and conduct a user study in order to develop two novel metrics for determining success based on personal preference. This is a challenging problem and we find that 50% of our model's predictions are likely to be found acceptable by at least 20% of people., Comment: 19 pages, 10 figures

Although end-to-end robot learning has shown some success for robot manipulation, the learned policies are often not sufficiently robust to variations in object pose or geometry. To improve the policy generalization, we introduce spatially-grounded parameterized motion primitives in our method HACMan++. Specifically, we propose an action representation consisting of three components: what primitive type (such as grasp or push) to execute, where the primitive will be grounded (e.g. where the gripper will make contact with the world), and how the primitive motion is executed, such as parameters specifying the push direction or grasp orientation. These three components define a novel discrete-continuous action space for reinforcement learning. Our framework enables robot agents to learn to chain diverse motion primitives together and select appropriate primitive parameters to complete long-horizon manipulation tasks. By grounding the primitives on a spatial location in the environment, our method is able to effectively generalize across object shape and pose variations. Our approach significantly outperforms existing methods, particularly in complex scenarios demanding both high-level sequential reasoning and object generalization. With zero-shot sim-to-real transfer, our policy succeeds in challenging real-world manipulation tasks, with generalization to unseen objects. Videos can be found on the project website: https://sgmp-rss2024.github.io.

In order to develop robots that can effectively serve as versatile and capable home assistants, it is crucial for them to reliably perceive and interact with a wide variety of objects across diverse environments. To this end, we proposed Open Vocabulary Mobile Manipulation as a key benchmark task for robotics: finding any object in a novel environment and placing it on any receptacle surface within that environment. We organized a NeurIPS 2023 competition featuring both simulation and real-world components to evaluate solutions to this task. Our baselines on the most challenging version of this task, using real perception in simulation, achieved only an 0.8% success rate; by the end of the competition, the best participants achieved an 10.8\% success rate, a 13x improvement. We observed that the most successful teams employed a variety of methods, yet two common threads emerged among the best solutions: enhancing error detection and recovery, and improving the integration of perception with decision-making processes. In this paper, we detail the results and methodologies used, both in simulation and real-world settings. We discuss the lessons learned and their implications for future research. Additionally, we compare performance in real and simulated environments, emphasizing the necessity for robust generalization to novel settings.

Agents that assist people need to have well-initialized policies that can adapt quickly to align with their partners' reward functions. Initializing policies to maximize performance with unknown partners can be achieved by bootstrapping nonlinear models using imitation learning over large, offline datasets. Such policies can require prohibitive computation to fine-tune in-situ and therefore may miss critical run-time information about a partner's reward function as expressed through their immediate behavior. In contrast, online logistic regression using low-capacity models performs rapid inference and fine-tuning updates and thus can make effective use of immediate in-task behavior for reward function alignment. However, these low-capacity models cannot be bootstrapped as effectively by offline datasets and thus have poor initializations. We propose BLR-HAC, Bootstrapped Logistic Regression for Human Agent Collaboration, which bootstraps large nonlinear models to learn the parameters of a low-capacity model which then uses online logistic regression for updates during collaboration. We test BLR-HAC in a simulated surface rearrangement task and demonstrate that it achieves higher zero-shot accuracy than shallow methods and takes far less computation to adapt online while still achieving similar performance to fine-tuned, large nonlinear models. For code, please see our project page https://sites.google.com/view/blr-hac., Comment: 10 pages, 4 figures, Accepted to AAMAS 2024

Recent developments in foundation models, like Large Language Models (LLMs) and Vision-Language Models (VLMs), trained on extensive data, facilitate flexible application across different tasks and modalities. Their impact spans various fields, including healthcare, education, and robotics. This paper provides an overview of the practical application of foundation models in real-world robotics, with a primary emphasis on the replacement of specific components within existing robot systems. The summary encompasses the perspective of input-output relationships in foundation models, as well as their role in perception, motion planning, and control within the field of robotics. This paper concludes with a discussion of future challenges and implications for practical robot applications.

Remarkable progress has been made in recent years in the fields of vision, language, and robotics. We now have vision models capable of recognizing objects based on language queries, navigation systems that can effectively control mobile systems, and grasping models that can handle a wide range of objects. Despite these advancements, general-purpose applications of robotics still lag behind, even though they rely on these fundamental capabilities of recognition, navigation, and grasping. In this paper, we adopt a systems-first approach to develop a new Open Knowledge-based robotics framework called OK-Robot. By combining Vision-Language Models (VLMs) for object detection, navigation primitives for movement, and grasping primitives for object manipulation, OK-Robot offers a integrated solution for pick-and-drop operations without requiring any training. To evaluate its performance, we run OK-Robot in 10 real-world home environments. The results demonstrate that OK-Robot achieves a 58.5% success rate in open-ended pick-and-drop tasks, representing a new state-of-the-art in Open Vocabulary Mobile Manipulation (OVMM) with nearly 1.8x the performance of prior work. On cleaner, uncluttered environments, OK-Robot's performance increases to 82%. However, the most important insight gained from OK-Robot is the critical role of nuanced details when combining Open Knowledge systems like VLMs with robotic modules. Videos of our experiments and code are available on our website: https://ok-robot.github.io, Comment: Github repo: https://github.com/ok-robot/ok-robot

In deployment scenarios such as homes and warehouses, mobile robots are expected to autonomously navigate for extended periods, seamlessly executing tasks articulated in terms that are intuitively understandable by human operators. We present GO To Any Thing (GOAT), a universal navigation system capable of tackling these requirements with three key features: a) Multimodal: it can tackle goals specified via category labels, target images, and language descriptions, b) Lifelong: it benefits from its past experience in the same environment, and c) Platform Agnostic: it can be quickly deployed on robots with different embodiments. GOAT is made possible through a modular system design and a continually augmented instance-aware semantic memory that keeps track of the appearance of objects from different viewpoints in addition to category-level semantics. This enables GOAT to distinguish between different instances of the same category to enable navigation to targets specified by images and language descriptions. In experimental comparisons spanning over 90 hours in 9 different homes consisting of 675 goals selected across 200+ different object instances, we find GOAT achieves an overall success rate of 83%, surpassing previous methods and ablations by 32% (absolute improvement). GOAT improves with experience in the environment, from a 60% success rate at the first goal to a 90% success after exploration. In addition, we demonstrate that GOAT can readily be applied to downstream tasks such as pick and place and social navigation.

HomeRobot (noun): An affordable compliant robot that navigates homes and manipulates a wide range of objects in order to complete everyday tasks. Open-Vocabulary Mobile Manipulation (OVMM) is the problem of picking any object in any unseen environment, and placing it in a commanded location. This is a foundational challenge for robots to be useful assistants in human environments, because it involves tackling sub-problems from across robotics: perception, language understanding, navigation, and manipulation are all essential to OVMM. In addition, integration of the solutions to these sub-problems poses its own substantial challenges. To drive research in this area, we introduce the HomeRobot OVMM benchmark, where an agent navigates household environments to grasp novel objects and place them on target receptacles. HomeRobot has two components: a simulation component, which uses a large and diverse curated object set in new, high-quality multi-room home environments; and a real-world component, providing a software stack for the low-cost Hello Robot Stretch to encourage replication of real-world experiments across labs. We implement both reinforcement learning and heuristic (model-based) baselines and show evidence of sim-to-real transfer. Our baselines achieve a 20% success rate in the real world; our experiments identify ways future research work improve performance. See videos on our website: https://ovmm.github.io/., Comment: 37 pages, 22 figures, 8 tables

Robots in home environments need to be able to learn new skills continuously as data becomes available, becoming ever more capable over time while using as little real-world data as possible. However, traditional robot learning approaches typically assume large amounts of iid data, which is inconsistent with this goal. In contrast, continual learning methods like CLEAR and SANE allow autonomous agents to learn off of a stream of non-iid samples; they, however, have not previously been demonstrated on real robotics platforms. In this work, we show how continual learning methods can be adapted for use on a real, low-cost home robot, and in particular look at the case where we have extremely small numbers of examples, in a task-id-free setting. Specifically, we propose SANER, a method for continuously learning a library of skills, and ABIP (Attention-Based Interaction Policies) as the backbone to support it. We learn four sequential kitchen tasks on a low-cost home robot, using only a handful of demonstrations per task.

Manipulating objects without grasping them is an essential component of human dexterity, referred to as non-prehensile manipulation. Non-prehensile manipulation may enable more complex interactions with the objects, but also presents challenges in reasoning about gripper-object interactions. In this work, we introduce Hybrid Actor-Critic Maps for Manipulation (HACMan), a reinforcement learning approach for 6D non-prehensile manipulation of objects using point cloud observations. HACMan proposes a temporally-abstracted and spatially-grounded object-centric action representation that consists of selecting a contact location from the object point cloud and a set of motion parameters describing how the robot will move after making contact. We modify an existing off-policy RL algorithm to learn in this hybrid discrete-continuous action representation. We evaluate HACMan on a 6D object pose alignment task in both simulation and in the real world. On the hardest version of our task, with randomized initial poses, randomized 6D goals, and diverse object categories, our policy demonstrates strong generalization to unseen object categories without a performance drop, achieving an 89% success rate on unseen objects in simulation and 50% success rate with zero-shot transfer in the real world. Compared to alternative action representations, HACMan achieves a success rate more than three times higher than the best baseline. With zero-shot sim2real transfer, our policy can successfully manipulate unseen objects in the real world for challenging non-planar goals, using dynamic and contact-rich non-prehensile skills. Videos can be found on the project website: https://hacman-2023.github.io., Comment: 7th Conference on Robot Learning (CoRL 2023)

In order for robots to follow open-ended instructions like "go open the brown cabinet over the sink", they require an understanding of both the scene geometry and the semantics of their environment. Robotic systems often handle these through separate pipelines, sometimes using very different representation spaces, which can be suboptimal when the two objectives conflict. In this work, we present USA-Net, a simple method for constructing a world representation that encodes both the semantics and spatial affordances of a scene in a differentiable map. This allows us to build a gradient-based planner which can navigate to locations in the scene specified using open-ended vocabulary. We use this planner to consistently generate trajectories which are both shorter 5-10% shorter and 10-30% closer to our goal query in CLIP embedding space than paths from comparable grid-based planners which don't leverage gradient information. To our knowledge, this is the first end-to-end differentiable planner optimizes for both semantics and affordance in a single implicit map. Code and visuals are available at our website: https://usa.bolte.cc/

Despite great strides in language-guided manipulation, existing work has been constrained to table-top settings. Table-tops allow for perfect and consistent camera angles, properties are that do not hold in mobile manipulation. Task plans that involve moving around the environment must be robust to egocentric views and changes in the plane and angle of grasp. A further challenge is ensuring this is all true while still being able to learn skills efficiently from limited data. We propose Spatial-Language Attention Policies (SLAP) as a solution. SLAP uses three-dimensional tokens as the input representation to train a single multi-task, language-conditioned action prediction policy. Our method shows an 80% success rate in the real world across eight tasks with a single model, and a 47.5% success rate when unseen clutter and unseen object configurations are introduced, even with only a handful of examples per task. This represents an improvement of 30% over prior work (20% given unseen distractors and configurations). We see a 4x improvement over baseline in mobile manipulation setting. In addition, we show how SLAPs robustness allows us to execute Task Plans from open-vocabulary instructions using a large language model for multi-step mobile manipulation. For videos, see the website: https://robotslap.github.io

Images are a convenient way to specify which particular object instance an embodied agent should navigate to. Solving this task requires semantic visual reasoning and exploration of unknown environments. We present a system that can perform this task in both simulation and the real world. Our modular method solves sub-tasks of exploration, goal instance re-identification, goal localization, and local navigation. We re-identify the goal instance in egocentric vision using feature-matching and localize the goal instance by projecting matched features to a map. Each sub-task is solved using off-the-shelf components requiring zero fine-tuning. On the HM3D InstanceImageNav benchmark, this system outperforms a baseline end-to-end RL policy 7x and a state-of-the-art ImageNav model 2.3x (56% vs 25% success). We deploy this system to a mobile robot platform and demonstrate effective real-world performance, achieving an 88% success rate across a home and an office environment.

Multi-object rearrangement is a crucial skill for service robots, and commonsense reasoning is frequently needed in this process. However, achieving commonsense arrangements requires knowledge about objects, which is hard to transfer to robots. Large language models (LLMs) are one potential source of this knowledge, but they do not naively capture information about plausible physical arrangements of the world. We propose LLM-GROP, which uses prompting to extract commonsense knowledge about semantically valid object configurations from an LLM and instantiates them with a task and motion planner in order to generalize to varying scene geometry. LLM-GROP allows us to go from natural-language commands to human-aligned object rearrangement in varied environments. Based on human evaluations, our approach achieves the highest rating while outperforming competitive baselines in terms of success rate while maintaining comparable cumulative action costs. Finally, we demonstrate a practical implementation of LLM-GROP on a mobile manipulator in real-world scenarios. Supplementary materials are available at: https://sites.google.com/view/llm-grop, Comment: Accpted by IEEE IROS 2023

Robots operating in human environments must be able to rearrange objects into semantically-meaningful configurations, even if these objects are previously unseen. In this work, we focus on the problem of building physically-valid structures without step-by-step instructions. We propose StructDiffusion, which combines a diffusion model and an object-centric transformer to construct structures given partial-view point clouds and high-level language goals, such as "set the table". Our method can perform multiple challenging language-conditioned multi-step 3D planning tasks using one model. StructDiffusion even improves the success rate of assembling physically-valid structures out of unseen objects by on average 16% over an existing multi-modal transformer model trained on specific structures. We show experiments on held-out objects in both simulation and on real-world rearrangement tasks. Importantly, we show how integrating both a diffusion model and a collision-discriminator model allows for improved generalization over other methods when rearranging previously-unseen objects. For videos and additional results, see our website: https://structdiffusion.github.io/., Comment: Accepted to Robotics: Science and Systems (RSS) 2023. The previous version appeared in CoRL Workshop on Language and Robot Learning 2022

We propose CLIP-Fields, an implicit scene model that can be used for a variety of tasks, such as segmentation, instance identification, semantic search over space, and view localization. CLIP-Fields learns a mapping from spatial locations to semantic embedding vectors. Importantly, we show that this mapping can be trained with supervision coming only from web-image and web-text trained models such as CLIP, Detic, and Sentence-BERT; and thus uses no direct human supervision. When compared to baselines like Mask-RCNN, our method outperforms on few-shot instance identification or semantic segmentation on the HM3D dataset with only a fraction of the examples. Finally, we show that using CLIP-Fields as a scene memory, robots can perform semantic navigation in real-world environments. Our code and demonstration videos are available here: https://mahis.life/clip-fields, Comment: Code, video, and interactive demonstrations available at https://mahis.life/clip-fields. Accepted for publication at Robotics: Science and Systems 2023 in Daegu, Korea

We introduce a new simulation benchmark "HandoverSim" for human-to-robot object handovers. To simulate the giver's motion, we leverage a recent motion capture dataset of hand grasping of objects. We create training and evaluation environments for the receiver with standardized protocols and metrics. We analyze the performance of a set of baselines and show a correlation with a real-world evaluation. Code is open sourced at https://handover-sim.github.io., Comment: Accepted to ICRA 2022

When humans design cost or goal specifications for robots, they often produce specifications that are ambiguous, underspecified, or beyond planners' ability to solve. In these cases, corrections provide a valuable tool for human-in-the-loop robot control. Corrections might take the form of new goal specifications, new constraints (e.g. to avoid specific objects), or hints for planning algorithms (e.g. to visit specific waypoints). Existing correction methods (e.g. using a joystick or direct manipulation of an end effector) require full teleoperation or real-time interaction. In this paper, we explore natural language as an expressive and flexible tool for robot correction. We describe how to map from natural language sentences to transformations of cost functions. We show that these transformations enable users to correct goals, update robot motions to accommodate additional user preferences, and recover from planning errors. These corrections can be leveraged to get 81% and 93% success rates on tasks where the original planner failed, with either one or two language corrections. Our method makes it possible to compose multiple constraints and generalizes to unseen scenes, objects, and sentences in simulated environments and real-world environments., Comment: 10 pages, 13 figures

Human-robot handover is a fundamental yet challenging task in human-robot interaction and collaboration. Recently, remarkable progressions have been made in human-to-robot handovers of unknown objects by using learning-based grasp generators. However, how to responsively generate smooth motions to take an object from a human is still an open question. Specifically, planning motions that take human comfort into account is not a part of the human-robot handover process in most prior works. In this paper, we propose to generate smooth motions via an efficient model-predictive control (MPC) framework that integrates perception and complex domain-specific constraints into the optimization problem. We introduce a learning-based grasp reachability model to select candidate grasps which maximize the robot's manipulability, giving it more freedom to satisfy these constraints. Finally, we integrate a neural net force/torque classifier that detects contact events from noisy data. We conducted human-to-robot handover experiments on a diverse set of objects with several users (N=4) and performed a systematic evaluation of each module. The study shows that the users preferred our MPC approach over the baseline system by a large margin. More results and videos are available at https://sites.google.com/nvidia.com/mpc-for-handover., Comment: Accepted to ICRA 2022

Rearrangement tasks have been identified as a crucial challenge for intelligent robotic manipulation, but few methods allow for precise construction of unseen structures. We propose a visual foresight model for pick-and-place rearrangement manipulation which is able to learn efficiently. In addition, we develop a multi-modal action proposal module which builds on the Goal-Conditioned Transporter Network, a state-of-the-art imitation learning method. Our image-based task planning method, Transporters with Visual Foresight, is able to learn from only a handful of data and generalize to multiple unseen tasks in a zero-shot manner. TVF is able to improve the performance of a state-of-the-art imitation learning method on unseen tasks in simulation and real robot experiments. In particular, the average success rate on unseen tasks improves from 55.4% to 78.5% in simulation experiments and from 30% to 63.3% in real robot experiments when given only tens of expert demonstrations. Video and code are available on our project website: https://chirikjianlab.github.io/tvf/, Comment: IEEE IROS 2022

Language model (LM) pre-training is useful in many language processing tasks. But can pre-trained LMs be further leveraged for more general machine learning problems? We propose an approach for using LMs to scaffold learning and generalization in general sequential decision-making problems. In this approach, goals and observations are represented as a sequence of embeddings, and a policy network initialized with a pre-trained LM predicts the next action. We demonstrate that this framework enables effective combinatorial generalization across different environments and supervisory modalities. We begin by assuming access to a set of expert demonstrations, and show that initializing policies with LMs and fine-tuning them via behavior cloning improves task completion rates by 43.6% in the VirtualHome environment. Next, we integrate an active data gathering procedure in which agents iteratively interact with the environment, relabel past "failed" experiences with new goals, and update their policies in a self-supervised loop. Active data gathering further improves combinatorial generalization, outperforming the best baseline by 25.1%. Finally, we explain these results by investigating three possible factors underlying the effectiveness of the LM-based policy. We find that sequential input representations (vs. fixed-dimensional feature vectors) and LM-based weight initialization are both important for generalization. Surprisingly, however, the format of the policy inputs encoding (e.g. as a natural language string vs. an arbitrary sequential encoding) has little influence. Together, these results suggest that language modeling induces representations that are useful for modeling not just language, but also goals and plans; these representations can aid learning and generalization even outside of language processing.

Accurate object rearrangement from vision is a crucial problem for a wide variety of real-world robotics applications in unstructured environments. We propose IFOR, Iterative Flow Minimization for Robotic Object Rearrangement, an end-to-end method for the challenging problem of object rearrangement for unknown objects given an RGBD image of the original and final scenes. First, we learn an optical flow model based on RAFT to estimate the relative transformation of the objects purely from synthetic data. This flow is then used in an iterative minimization algorithm to achieve accurate positioning of previously unseen objects. Crucially, we show that our method applies to cluttered scenes, and in the real world, while training only on synthetic data. Videos are available at https://imankgoyal.github.io/ifor.html.

We have recently introduced a system that automatically generates robotic planning operators from human demonstrations. One feature of our system is the operator count, which keeps track of the application frequency of every operator within the demonstrations. In this extended abstract, we show that we can use the count to slim down domains with the goal of decreasing the search time for long-horizon planning goals. The conceptual idea behind our approach is that we would like to prioritize operators that have occurred more often in the demonstrations over those that were not observed so frequently. We, therefore, propose to limit the domain only to the most popular operators. If this subset of operators is not sufficient to find a plan, we iteratively expand this subset of operators. We show that this significantly reduces the search time for long-horizon planning goals., Comment: Accepted and presented as extended abstract, 5th Workshop on Semantic Policy and Action Representations for Autonomous Robots (SPAR), Prague, CzechRepublic, Online, 2021, at IROS 2021

When robots operate in human environments, it's critical that humans can quickly teach them new concepts: object-centric properties of the environment that they care about (e.g. objects near, upright, etc). However, teaching a new perceptual concept from high-dimensional robot sensor data (e.g. point clouds) is demanding, requiring an unrealistic amount of human labels. To address this, we propose a framework called Perceptual Concept Bootstrapping (PCB). First, we leverage the inherently lower-dimensional privileged information, e.g., object poses and bounding boxes, available from a simulator only at training time to rapidly learn a low-dimensional, geometric concept from minimal human input. Second, we treat this low-dimensional concept as an automatic labeler to synthesize a large-scale high-dimensional data set with the simulator. With these two key ideas, PCB alleviates human label burden while still learning perceptual concepts that work with real sensor input where no privileged information is available. We evaluate PCB for learning spatial concepts that describe object state or multi-object relationships, and show it achieves superior performance compared to baseline methods. We also demonstrate the utility of the learned concepts in motion planning tasks on a 7-DoF Franka Panda robot., Comment: 9 pages, 10 figures

Geometric organization of objects into semantically meaningful arrangements pervades the built world. As such, assistive robots operating in warehouses, offices, and homes would greatly benefit from the ability to recognize and rearrange objects into these semantically meaningful structures. To be useful, these robots must contend with previously unseen objects and receive instructions without significant programming. While previous works have examined recognizing pairwise semantic relations and sequential manipulation to change these simple relations none have shown the ability to arrange objects into complex structures such as circles or table settings. To address this problem we propose a novel transformer-based neural network, StructFormer, which takes as input a partial-view point cloud of the current object arrangement and a structured language command encoding the desired object configuration. We show through rigorous experiments that StructFormer enables a physical robot to rearrange novel objects into semantically meaningful structures with multi-object relational constraints inferred from the language command.

Sequential manipulation tasks require a robot to perceive the state of an environment and plan a sequence of actions leading to a desired goal state. In such tasks, the ability to reason about spatial relations among object entities from raw sensor inputs is crucial in order to determine when a task has been completed and which actions can be executed. In this work, we propose SORNet (Spatial Object-Centric Representation Network), a framework for learning object-centric representations from RGB images conditioned on a set of object queries, represented as image patches called canonical object views. With only a single canonical view per object and no annotation, SORNet generalizes zero-shot to object entities whose shape and texture are both unseen during training. We evaluate SORNet on various spatial reasoning tasks such as spatial relation classification and relative direction regression in complex tabletop manipulation scenarios and show that SORNet significantly outperforms baselines including state-of-the-art representation learning techniques. We also demonstrate the application of the representation learned by SORNet on visual-servoing and task planning for sequential manipulation on a real robot., Comment: CoRL 2021 Best Systems Paper Finalist; Code and data available at https://github.com/wentaoyuan/sornet

Human environments contain numerous objects configured in a variety of arrangements. Our goal is to enable robots to repose previously unseen objects according to learned semantic relationships in novel environments. We break this problem down into two parts: (1) finding physically valid locations for the objects and (2) determining if those poses satisfy learned, high-level semantic relationships. We build our models and training from the ground up to be tightly integrated with our proposed planning algorithm for semantic placement of unknown objects. We train our models purely in simulation, with no fine-tuning needed for use in the real world. Our approach enables motion planning for semantic rearrangement of unknown objects in scenes with varying geometry from only RGB-D sensing. Our experiments through a set of simulated ablations demonstrate that using a relational classifier alone is not sufficient for reliable planning. We further demonstrate the ability of our planner to generate and execute diverse manipulation plans through a set of real-world experiments with a variety of objects.

Seemingly simple natural language requests to a robot are generally underspecified, for example "Can you bring me the wireless mouse?" Flat images of candidate mice may not provide the discriminative information needed for "wireless." The world, and objects in it, are not flat images but complex 3D shapes. If a human requests an object based on any of its basic properties, such as color, shape, or texture, robots should perform the necessary exploration to accomplish the task. In particular, while substantial effort and progress has been made on understanding explicitly visual attributes like color and category, comparatively little progress has been made on understanding language about shapes and contours. In this work, we introduce a novel reasoning task that targets both visual and non-visual language about 3D objects. Our new benchmark, ShapeNet Annotated with Referring Expressions (SNARE) requires a model to choose which of two objects is being referenced by a natural language description. We introduce several CLIP-based models for distinguishing objects and demonstrate that while recent advances in jointly modeling vision and language are useful for robotic language understanding, it is still the case that these image-based models are weaker at understanding the 3D nature of objects -- properties which play a key role in manipulation. We find that adding view estimation to language grounding models improves accuracy on both SNARE and when identifying objects referred to in language on a robot platform, but note that a large gap remains between these models and human performance., Comment: Conference on Robot Learning (CoRL) 2021

Natural language provides an accessible and expressive interface to specify long-term tasks for robotic agents. However, non-experts are likely to specify such tasks with high-level instructions, which abstract over specific robot actions through several layers of abstraction. We propose that key to bridging this gap between language and robot actions over long execution horizons are persistent representations. We propose a persistent spatial semantic representation method, and show how it enables building an agent that performs hierarchical reasoning to effectively execute long-term tasks. We evaluate our approach on the ALFRED benchmark and achieve state-of-the-art results, despite completely avoiding the commonly used step-by-step instructions., Comment: Presented at CoRL 2021

Robots will be expected to manipulate a wide variety of objects in complex and arbitrary ways as they become more widely used in human environments. As such, the rearrangement of objects has been noted to be an important benchmark for AI capabilities in recent years. We propose NeRP (Neural Rearrangement Planning), a deep learning based approach for multi-step neural object rearrangement planning which works with never-before-seen objects, that is trained on simulation data, and generalizes to the real world. We compare NeRP to several naive and model-based baselines, demonstrating that our approach is measurably better and can efficiently arrange unseen objects in fewer steps and with less planning time. Finally, we demonstrate it on several challenging rearrangement problems in the real world., Comment: Please refer to our supplementary video: https://youtu.be/CJb1IzH94eo

Automated planning enables robots to find plans to achieve complex, long-horizon tasks, given a planning domain. This planning domain consists of a list of actions, with their associated preconditions and effects, and is usually manually defined by a human expert, which is very time-consuming or even infeasible. In this paper, we introduce a novel method for generating this domain automatically from human demonstrations. First, we automatically segment and recognize the different observed actions from human demonstrations. From these demonstrations, the relevant preconditions and effects are obtained, and the associated planning operators are generated. Finally, a sequence of actions that satisfies a user-defined goal can be planned using a symbolic planner. The generated plan is executed in a simulated environment by the TIAGo robot. We tested our method on a dataset of 12 demonstrations collected from three different participants. The results show that our method is able to generate executable plans from using one single demonstration with a 92% success rate, and 100% when the information from all demonstrations are included, even for previously unknown stacking goals., Comment: 8 pages, 5 figures

In many applications, including logistics and manufacturing, robot manipulators operate in semi-structured environments alongside humans or other robots. These environments are largely static, but they may contain some movable obstacles that the robot must avoid. Manipulation tasks in these applications are often highly repetitive, but require fast and reliable motion planning capabilities, often under strict time constraints. Existing preprocessing-based approaches are beneficial when the environments are highly-structured, but their performance degrades in the presence of movable obstacles, since these are not modelled a priori. We propose a novel preprocessing-based method called Alternative Paths Planner (APP) that provides provably fixed-time planning guarantees in semi-structured environments. APP plans a set of alternative paths offline such that, for any configuration of the movable obstacles, at least one of the paths from this set is collision-free. During online execution, a collision-free path can be looked up efficiently within a few microseconds. We evaluate APP on a 7 DoF robot arm in semi-structured domains of varying complexity and demonstrate that APP is several orders of magnitude faster than state-of-the-art motion planners for each domain. We further validate this approach with real-time experiments on a robotic manipulator.

Human-robot object handovers have been an actively studied area of robotics over the past decade; however, very few techniques and systems have addressed the challenge of handing over diverse objects with arbitrary appearance, size, shape, and rigidity. In this paper, we present a vision-based system that enables reactive human-to-robot handovers of unknown objects. Our approach combines closed-loop motion planning with real-time, temporally-consistent grasp generation to ensure reactivity and motion smoothness. Our system is robust to different object positions and orientations, and can grasp both rigid and non-rigid objects. We demonstrate the generalizability, usability, and robustness of our approach on a novel benchmark set of 26 diverse household objects, a user study with naive users (N=6) handing over a subset of 15 objects, and a systematic evaluation examining different ways of handing objects. More results and videos can be found at https://sites.google.com/nvidia.com/handovers-of-arbitrary-objects., Comment: Accepted to the International Conference on Robotics and Automation (ICRA) 2021

Zero-shot execution of unseen robotic tasks is important to allowing robots to perform a wide variety of tasks in human environments, but collecting the amounts of data necessary to train end-to-end policies in the real-world is often infeasible. We describe an approach for sim-to-real training that can accomplish unseen robotic tasks using models learned in simulation to ground components of a simple task planner. We learn a library of parameterized skills, along with a set of predicates-based preconditions and termination conditions, entirely in simulation. We explore a block-stacking task because it has a clear structure, where multiple skills must be chained together, but our methods are applicable to a wide range of other problems and domains, and can transfer from simulation to the real-world with no fine tuning. The system is able to recognize failures and accomplish long-horizon tasks from perceptual input, which is critical for real-world execution. We evaluate our proposed approach in both simulation and in the real-world, showing an increase in success rate from 91.6% to 98% in simulation and from 10% to 80% success rate in the real-world as compared with naive baselines. For experiment videos including both real-world and simulation, see: https://www.youtube.com/playlist?list=PL-oD0xHUngeLfQmpngYkGFZarstfPOXqX, Comment: IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) 2021

Transfer of objects between humans and robots is a critical capability for collaborative robots. Although there has been a recent surge of interest in human-robot handovers, most prior research focus on robot-to-human handovers. Further, work on the equally critical human-to-robot handovers often assumes humans can place the object in the robot's gripper. In this paper, we propose an approach for human-to-robot handovers in which the robot meets the human halfway, by classifying the human's grasp of the object and quickly planning a trajectory accordingly to take the object from the human's hand according to their intent. To do this, we collect a human grasp dataset which covers typical ways of holding objects with various hand shapes and poses, and learn a deep model on this dataset to classify the hand grasps into one of these categories. We present a planning and execution approach that takes the object from the human hand according to the detected grasp and hand position, and replans as necessary when the handover is interrupted. Through a systematic evaluation, we demonstrate that our system results in more fluent handovers versus two baselines. We also present findings from a user study (N = 9) demonstrating the effectiveness and usability of our approach with naive users in different scenarios. More results and videos can be found at http://wyang.me/handovers.

While robots can learn models to solve many manipulation tasks from raw visual input, they cannot usually use these models to solve new problems. On the other hand, symbolic planning methods such as STRIPS have long been able to solve new problems given only a domain definition and a symbolic goal, but these approaches often struggle on the real world robotic tasks due to the challenges of grounding these symbols from sensor data in a partially-observable world. We propose Deep Planning Domain Learning (DPDL), an approach that combines the strengths of both methods to learn a hierarchical model. DPDL learns a high-level model which predicts values for a large set of logical predicates consisting of the current symbolic world state, and separately learns a low-level policy which translates symbolic operators into executable actions on the robot. This allows us to perform complex, multi-step tasks even when the robot has not been explicitly trained on them. We show our method on manipulation tasks in a photorealistic kitchen scenario., Comment: 7 pages, 6 figures. Conference paper accepted in International conference on Robotics and Automation (ICRA) 2020

Grasping in cluttered environments is a fundamental but challenging robotic skill. It requires both reasoning about unseen object parts and potential collisions with the manipulator. Most existing data-driven approaches avoid this problem by limiting themselves to top-down planar grasps which is insufficient for many real-world scenarios and greatly limits possible grasps. We present a method that plans 6-DOF grasps for any desired object in a cluttered scene from partial point cloud observations. Our method achieves a grasp success of 80.3%, outperforming baseline approaches by 17.6% and clearing 9 cluttered table scenes (which contain 23 unknown objects and 51 picks in total) on a real robotic platform. By using our learned collision checking module, we can even reason about effective grasp sequences to retrieve objects that are not immediately accessible. Supplementary video can be found at https://youtu.be/w0B5S-gCsJk., Comment: Accepted to the International Conference on Robotics and Automation (ICRA) 2020

We address goal-based imitation learning, where the aim is to output the symbolic goal from a third-person video demonstration. This enables the robot to plan for execution and reproduce the same goal in a completely different environment. The key challenge is that the goal of a video demonstration is often ambiguous at the level of semantic actions. The human demonstrators might unintentionally achieve certain subgoals in the demonstrations with their actions. Our main contribution is to propose a motion reasoning framework that combines task and motion planning to disambiguate the true intention of the demonstrator in the video demonstration. This allows us to robustly recognize the goals that cannot be disambiguated by previous action-based approaches. We evaluate our approach by collecting a dataset of 96 video demonstrations in a mockup kitchen environment. We show that our motion reasoning plays an important role in recognizing the actual goal of the demonstrator and improves the success rate by over 20%. We further show that by using the automatically inferred goal from the video demonstration, our robot is able to reproduce the same task in a real kitchen environment.

To solve multi-step manipulation tasks in the real world, an autonomous robot must take actions to observe its environment and react to unexpected observations. This may require opening a drawer to observe its contents or moving an object out of the way to examine the space behind it. Upon receiving a new observation, the robot must update its belief about the world and compute a new plan of action. In this work, we present an online planning and execution system for robots faced with these challenges. We perform deterministic cost-sensitive planning in the space of hybrid belief states to select likely-to-succeed observation actions and continuous control actions. After execution and observation, we replan using our new state estimate. We initially enforce that planner reuses the structure of the unexecuted tail of the last plan. This both improves planning efficiency and ensures that the overall policy does not undo its progress towards achieving the goal. Our approach is able to efficiently solve partially observable problems both in simulation and in a real-world kitchen., Comment: IEEE International Conference on Robotics and Automation (ICRA), 2020

Widespread adoption of self-driving cars will depend not only on their safety but largely on their ability to interact with human users. Just like human drivers, self-driving cars will be expected to understand and safely follow natural-language directions that suddenly alter the pre-planned route according to user's preference or in presence of ambiguities, particularly in locations with poor or outdated map coverage. To this end, we propose a language-grounded driving agent implementing a hierarchical policy using recurrent layers and gated attention. The hierarchical approach enables us to reason both in terms of high-level language instructions describing long time horizons and low-level, complex, continuous state/action spaces required for real-time control of a self-driving car. We train our policy with conditional imitation learning from realistic language data collected from human drivers and navigators. Through quantitative and interactive experiments within the CARLA framework, we show that our model can successfully interpret language instructions and follow them safely, even when generalizing to previously unseen environments. Code and video are available at https://sites.google.com/view/language-grounded-driving., Comment: Accepted by the 3rd Conference on Robot Learning, Osaka, Japan (CoRL 2019)

Effective human-robot collaboration requires informed anticipation. The robot must anticipate the human's actions, but also react quickly and intuitively when its predictions are wrong. The robot must plan its actions to account for the human's own plan, with the knowledge that the human's behavior will change based on what the robot actually does. This cyclical game of predicting a human's future actions and generating a corresponding motion plan is extremely difficult to model using standard techniques. In this work, we describe a novel Model Predictive Control (MPC)-based framework for finding optimal trajectories in a collaborative, multi-agent setting, in which we simultaneously plan for the robot while predicting the actions of its external collaborators. We use human-robot handovers to demonstrate that with a strong model of the collaborator, our framework produces fluid, reactive human-robot interactions in novel, cluttered environments. Our method efficiently generates coordinated trajectories, and achieves a high success rate in handover, even in the presence of significant sensor noise., Comment: 3 figures, 7 pages

Current Reinforcement Learning (RL) algorithms struggle with long-horizon tasks where time can be wasted exploring dead ends and task progress may be easily reversed. We develop the SPOT framework, which explores within action safety zones, learns about unsafe regions without exploring them, and prioritizes experiences that reverse earlier progress to learn with remarkable efficiency. The SPOT framework successfully completes simulated trials of a variety of tasks, improving a baseline trial success rate from 13% to 100% when stacking 4 cubes, from 13% to 99% when creating rows of 4 cubes, and from 84% to 95% when clearing toys arranged in adversarial patterns. Efficiency with respect to actions per trial typically improves by 30% or more, while training takes just 1-20k actions, depending on the task. Furthermore, we demonstrate direct sim to real transfer. We are able to create real stacks in 100% of trials with 61% efficiency and real rows in 100% of trials with 59% efficiency by directly loading the simulation-trained model on the real robot with no additional real-world fine-tuning. To our knowledge, this is the first instance of reinforcement learning with successful sim to real transfer applied to long term multi-step tasks such as block-stacking and row-making with consideration of progress reversal. Code is available at https://github.com/jhu-lcsr/good_robot ., Comment: Accepted to the journal IEEE Robotics and Automation Letters (RA-L) and to be presented at IROS 2020. This is a minor update to v3. 8 pages, 6 figures, 3 tables, 1 algorithm. Code is available at https://github.com/jhu-lcsr/good_robot and a video overview is at https://youtu.be/MbCuEZadkIw

It is difficult to create robust, reusable, and reactive behaviors for robots that can be easily extended and combined. Frameworks such as Behavior Trees are flexible but difficult to characterize, especially when designing reactions and recovery behaviors to consistently converge to a desired goal condition. We propose a framework which we call Robust Logical-Dynamical Systems (RLDS), which combines the advantages of task representations like behavior trees with theoretical guarantees on performance. RLDS can also be constructed automatically from simple sequential task plans and will still achieve robust, reactive behavior in dynamic real-world environments. In this work, we describe both our proposed framework and a case study on a simple household manipulation task, with examples for how specific pieces can be implemented to achieve robust behavior. Finally, we show how in the context of these manipulation tasks, a combination of an RLDS with planning can achieve better results under adversarial conditions., Comment: 9 pages, extended version of IROS 2019 paper

High-level human instructions often correspond to behaviors with multiple implicit steps. In order for robots to be useful in the real world, they must be able to to reason over both motions and intermediate goals implied by human instructions. In this work, we propose a framework for learning representations that convert from a natural-language command to a sequence of intermediate goals for execution on a robot. A key feature of this framework is prospection, training an agent not just to correctly execute the prescribed command, but to predict a horizon of consequences of an action before taking it. We demonstrate the fidelity of plans generated by our framework when interpreting real, crowd-sourced natural language commands for a robot in simulated scenes., Comment: Accepted to ICRA 2019; extended version with appendix containing additional results

How can we enable users to create effective, perception-driven task plans for collaborative robots? We conducted a 35-person user study with the Behavior Tree-based CoSTAR system to determine which strategies for end user creation of generalizable robot task plans are most usable and effective. CoSTAR allows domain experts to author complex, perceptually grounded task plans for collaborative robots. As a part of CoSTAR's wide range of capabilities, it allows users to specify SmartMoves: abstract goals such as "pick up component A from the right side of the table." Users were asked to perform pick-and-place assembly tasks with either SmartMoves or one of three simpler baseline versions of CoSTAR. Overall, participants found CoSTAR to be highly usable, with an average System Usability Scale score of 73.4 out of 100. SmartMove also helped users perform tasks faster and more effectively; all SmartMove users completed the first two tasks, while not all users completed the tasks using the other strategies. SmartMove users showed better performance for incorporating perception across all three tasks., Comment: 7 pages; IROS 2018

A robot can now grasp an object more effectively than ever before, but once it has the object what happens next? We show that a mild relaxation of the task and workspace constraints implicit in existing object grasping datasets can cause neural network based grasping algorithms to fail on even a simple block stacking task when executed under more realistic circumstances. To address this, we introduce the JHU CoSTAR Block Stacking Dataset (BSD), where a robot interacts with 5.1 cm colored blocks to complete an order-fulfillment style block stacking task. It contains dynamic scenes and real time-series data in a less constrained environment than comparable datasets. There are nearly 12,000 stacking attempts and over 2 million frames of real data. We discuss the ways in which this dataset provides a valuable resource for a broad range of other topics of investigation. We find that hand-designed neural networks that work on prior datasets do not generalize to this task. Thus, to establish a baseline for this dataset, we demonstrate an automated search of neural network based models using a novel multiple-input HyperTree MetaModel, and find a final model which makes reasonable 3D pose predictions for grasping and stacking on our dataset. The CoSTAR BSD, code, and instructions are available at https://sites.google.com/site/costardataset., Comment: This is a major revision refocusing the topic towards the JHU CoSTAR Block Stacking Dataset, workspace constraints, and a comparison of HyperTrees with hand-designed algorithms. 12 pages, 10 figures, and 3 tables

Prospection, the act of predicting the consequences of many possible futures, is intrinsic to human planning and action, and may even be at the root of consciousness. Surprisingly, this idea has been explored comparatively little in robotics. In this work, we propose a neural network architecture and associated planning algorithm that (1) learns a representation of the world useful for generating prospective futures after the application of high-level actions, (2) uses this generative model to simulate the result of sequences of high-level actions in a variety of environments, and (3) uses this same representation to evaluate these actions and perform tree search to find a sequence of high-level actions in a new environment. Models are trained via imitation learning on a variety of domains, including navigation, pick-and-place, and a surgical robotics task. Our approach allows us to visualize intermediate motion goals and learn to plan complex activity from visual information., Comment: 8 pages, IEEE format, currently in review

Fast, collision-free motion through unknown environments remains a challenging problem for robotic systems. In these situations, the robot's ability to reason about its future motion is often severely limited by sensor field of view (FOV). By contrast, biological systems routinely make decisions by taking into consideration what might exist beyond their FOV based on prior experience. In this paper, we present an approach for predicting occupancy map representations of sensor data for future robot motions using deep neural networks. We evaluate several deep network architectures, including purely generative and adversarial models. Testing on both simulated and real environments we demonstrated performance both qualitatively and quantitatively, with SSIM similarity measure up to 0.899. We showed that it is possible to make predictions about occupied space beyond the physical robot's FOV from simulated training data. In the future, this method will allow robots to navigate through unknown environments in a faster, safer manner., Comment: 7 pages