Akshara Rai

Chenyu Yang, Guo Ning Sue, Zhongyu Li et al. · berkeley

This paper tackles the problem of robots collaboratively towing a load with cables to a specified goal location while avoiding collisions in real time. The introduction of cables (as opposed to rigid links) enables the robotic team to travel through narrow spaces by changing its intrinsic dimensions through slack/taut switches of the cable. However, this is a challenging problem because of the hybrid mode switches and the dynamical coupling among multiple robots and the load. Previous attempts at addressing such a problem were performed offline and do not consider avoiding obstacles online. In this paper, we introduce a cascaded planning scheme with a parallelized centralized trajectory optimization that deals with hybrid mode switches. We additionally develop a set of decentralized planners per robot, which enables our approach to solve the problem of collaborative load manipulation online. We develop and demonstrate one of the first collaborative autonomy framework that is able to move a cable-towed load, which is too heavy to move by a single robot, through narrow spaces with real-time feedback and reactive planning in experiments.

Vidhi Jain, Yixin Lin, Eric Undersander et al. · cmu

Every home is different, and every person likes things done in their particular way. Therefore, home robots of the future need to both reason about the sequential nature of day-to-day tasks and generalize to user's preferences. To this end, we propose a Transformer Task Planner(TTP) that learns high-level actions from demonstrations by leveraging object attribute-based representations. TTP can be pre-trained on multiple preferences and shows generalization to unseen preferences using a single demonstration as a prompt in a simulated dishwasher loading task. Further, we demonstrate real-world dish rearrangement using TTP with a Franka Panda robotic arm, prompted using a single human demonstration.

Xavier Puig, Eric Undersander, Andrew Szot et al.

We present Habitat 3.0: a simulation platform for studying collaborative human-robot tasks in home environments. Habitat 3.0 offers contributions across three dimensions: (1) Accurate humanoid simulation: addressing challenges in modeling complex deformable bodies and diversity in appearance and motion, all while ensuring high simulation speed. (2) Human-in-the-loop infrastructure: enabling real human interaction with simulated robots via mouse/keyboard or a VR interface, facilitating evaluation of robot policies with human input. (3) Collaborative tasks: studying two collaborative tasks, Social Navigation and Social Rearrangement. Social Navigation investigates a robot's ability to locate and follow humanoid avatars in unseen environments, whereas Social Rearrangement addresses collaboration between a humanoid and robot while rearranging a scene. These contributions allow us to study end-to-end learned and heuristic baselines for human-robot collaboration in-depth, as well as evaluate them with humans in the loop. Our experiments demonstrate that learned robot policies lead to efficient task completion when collaborating with unseen humanoid agents and human partners that might exhibit behaviors that the robot has not seen before. Additionally, we observe emergent behaviors during collaborative task execution, such as the robot yielding space when obstructing a humanoid agent, thereby allowing the effective completion of the task by the humanoid agent. Furthermore, our experiments using the human-in-the-loop tool demonstrate that our automated evaluation with humanoids can provide an indication of the relative ordering of different policies when evaluated with real human collaborators. Habitat 3.0 unlocks interesting new features in simulators for Embodied AI, and we hope it paves the way for a new frontier of embodied human-AI interaction capabilities.

So Yeon Min, Xavi Puig, Devendra Singh Chaplot et al. · cmu

Language is never spoken in a vacuum. It is expressed, comprehended, and contextualized within the holistic backdrop of the speaker's history, actions, and environment. Since humans are used to communicating efficiently with situated language, the practicality of robotic assistants hinge on their ability to understand and act upon implicit and situated instructions. In traditional instruction following paradigms, the agent acts alone in an empty house, leading to language use that is both simplified and artificially "complete." In contrast, we propose situated instruction following, which embraces the inherent underspecification and ambiguity of real-world communication with the physical presence of a human speaker. The meaning of situated instructions naturally unfold through the past actions and the expected future behaviors of the human involved. Specifically, within our settings we have instructions that (1) are ambiguously specified, (2) have temporally evolving intent, (3) can be interpreted more precisely with the agent's dynamic actions. Our experiments indicate that state-of-the-art Embodied Instruction Following (EIF) models lack holistic understanding of situated human intention.

Xiaoyu Huang, Dhruv Batra, Akshara Rai et al.

We present Skill Transformer, an approach for solving long-horizon robotic tasks by combining conditional sequence modeling and skill modularity. Conditioned on egocentric and proprioceptive observations of a robot, Skill Transformer is trained end-to-end to predict both a high-level skill (e.g., navigation, picking, placing), and a whole-body low-level action (e.g., base and arm motion), using a transformer architecture and demonstration trajectories that solve the full task. It retains the composability and modularity of the overall task through a skill predictor module while reasoning about low-level actions and avoiding hand-off errors, common in modular approaches. We test Skill Transformer on an embodied rearrangement benchmark and find it performs robust task planning and low-level control in new scenarios, achieving a 2.5x higher success rate than baselines in hard rearrangement problems.

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.

Shuxiao Chen, Bike Zhang, Mark W. Mueller et al.

Reinforcement learning (RL) has become a promising approach to developing controllers for quadrupedal robots. Conventionally, an RL design for locomotion follows a position-based paradigm, wherein an RL policy outputs target joint positions at a low frequency that are then tracked by a high-frequency proportional-derivative (PD) controller to produce joint torques. In contrast, for the model-based control of quadrupedal locomotion, there has been a paradigm shift from position-based control to torque-based control. In light of the recent advances in model-based control, we explore an alternative to the position-based RL paradigm, by introducing a torque-based RL framework, where an RL policy directly predicts joint torques at a high frequency, thus circumventing the use of a PD controller. The proposed learning torque control framework is validated with extensive experiments, in which a quadruped is capable of traversing various terrain and resisting external disturbances while following user-specified commands. Furthermore, compared to learning position control, learning torque control demonstrates the potential to achieve a higher reward and is more robust to significant external disturbances. To our knowledge, this is the first sim-to-real attempt for end-to-end learning torque control of quadrupedal locomotion.

Rishi Hazra, Brian Chen, Akshara Rai et al.

To enable progress towards egocentric agents capable of understanding everyday tasks specified in natural language, we propose a benchmark and a synthetic dataset called Egocentric Task Verification (EgoTV). The goal in EgoTV is to verify the execution of tasks from egocentric videos based on the natural language description of these tasks. EgoTV contains pairs of videos and their task descriptions for multi-step tasks -- these tasks contain multiple sub-task decompositions, state changes, object interactions, and sub-task ordering constraints. In addition, EgoTV also provides abstracted task descriptions that contain only partial details about ways to accomplish a task. Consequently, EgoTV requires causal, temporal, and compositional reasoning of video and language modalities, which is missing in existing datasets. We also find that existing vision-language models struggle at such all round reasoning needed for task verification in EgoTV. Inspired by the needs of EgoTV, we propose a novel Neuro-Symbolic Grounding (NSG) approach that leverages symbolic representations to capture the compositional and temporal structure of tasks. We demonstrate NSG's capability towards task tracking and verification on our EgoTV dataset and a real-world dataset derived from CrossTask (CTV). We open-source the EgoTV and CTV datasets and the NSG model for future research on egocentric assistive agents.

Matthew Chang, Gunjan Chhablani, Alexander Clegg et al.

We present a benchmark for Planning And Reasoning Tasks in humaN-Robot collaboration (PARTNR) designed to study human-robot coordination in household activities. PARTNR tasks exhibit characteristics of everyday tasks, such as spatial, temporal, and heterogeneous agent capability constraints. We employ a semi-automated task generation pipeline using Large Language Models (LLMs), incorporating simulation in the loop for grounding and verification. PARTNR stands as the largest benchmark of its kind, comprising 100,000 natural language tasks, spanning 60 houses and 5,819 unique objects. We analyze state-of-the-art LLMs on PARTNR tasks, across the axes of planning, perception and skill execution. The analysis reveals significant limitations in SoTA models, such as poor coordination and failures in task tracking and recovery from errors. When LLMs are paired with real humans, they require 1.5x as many steps as two humans collaborating and 1.1x more steps than a single human, underscoring the potential for improvement in these models. We further show that fine-tuning smaller LLMs with planning data can achieve performance on par with models 9 times larger, while being 8.6x faster at inference. Overall, PARTNR highlights significant challenges facing collaborative embodied agents and aims to drive research in this direction.

Maithili Patel, Xavier Puig, Ruta Desai et al.

Assistive agents should be able to perform under-specified long-horizon tasks while respecting user preferences. We introduce Actively Discovering and Adapting to Preferences for any Task (ADAPT) -- a benchmark designed to evaluate agents' ability to adhere to user preferences across various household tasks through active questioning. Next, we propose Reflection-DPO, a novel training approach for adapting large language models (LLMs) to the task of active questioning. Reflection-DPO finetunes a 'student' LLM to follow the actions of a privileged 'teacher' LLM, and optionally ask a question to gather necessary information to better predict the teacher action. We find that prior approaches that use state-of-the-art LLMs fail to sufficiently follow user preferences in ADAPT due to insufficient questioning and poor adherence to elicited preferences. In contrast, Reflection-DPO achieves a higher rate of satisfying user preferences, outperforming a zero-shot chain-of-thought baseline by 6.1% on unseen users.

Irmak Guzey, Haozhi Qi, Julen Urain et al. · cmu, meta-ai

Learning multi-fingered robot policies from humans performing daily tasks in natural environments has long been a grand goal in the robotics community. Achieving this would mark significant progress toward generalizable robot manipulation in human environments, as it would reduce the reliance on labor-intensive robot data collection. Despite substantial efforts, progress toward this goal has been bottle-necked by the embodiment gap between humans and robots, as well as by difficulties in extracting relevant contextual and motion cues that enable learning of autonomous policies from in-the-wild human videos. We claim that with simple yet sufficiently powerful hardware for obtaining human data and our proposed framework AINA, we are now one significant step closer to achieving this dream. AINA enables learning multi-fingered policies from data collected by anyone, anywhere, and in any environment using Aria Gen 2 glasses. These glasses are lightweight and portable, feature a high-resolution RGB camera, provide accurate on-board 3D head and hand poses, and offer a wide stereo view that can be leveraged for depth estimation of the scene. This setup enables the learning of 3D point-based policies for multi-fingered hands that are robust to background changes and can be deployed directly without requiring any robot data (including online corrections, reinforcement learning, or simulation). We compare our framework against prior human-to-robot policy learning approaches, ablate our design choices, and demonstrate results across nine everyday manipulation tasks. Robot rollouts are best viewed on our website: https://aina-robot.github.io.

Andrew Szot, Unnat Jain, Dhruv Batra et al.

We present the task of "Social Rearrangement", consisting of cooperative everyday tasks like setting up the dinner table, tidying a house or unpacking groceries in a simulated multi-agent environment. In Social Rearrangement, two robots coordinate to complete a long-horizon task, using onboard sensing and egocentric observations, and no privileged information about the environment. We study zero-shot coordination (ZSC) in this task, where an agent collaborates with a new partner, emulating a scenario where a robot collaborates with a new human partner. Prior ZSC approaches struggle to generalize in our complex and visually rich setting, and on further analysis, we find that they fail to generate diverse coordination behaviors at training time. To counter this, we propose Behavior Diversity Play (BDP), a novel ZSC approach that encourages diversity through a discriminability objective. Our results demonstrate that BDP learns adaptive agents that can tackle visual coordination, and zero-shot generalize to new partners in unseen environments, achieving 35% higher success and 32% higher efficiency compared to baselines.

Ayush Agrawal, Shuxiao Chen, Akshara Rai et al.

In this paper, we present a framework rooted in control and planning that enables quadrupedal robots to traverse challenging terrains with discrete footholds using visual feedback. Navigating discrete terrain is challenging for quadrupeds because the motion of the robot can be aperiodic, highly dynamic, and blind for the hind legs of the robot. Additionally, the robot needs to reason over both the feasible footholds as well as robot velocity by speeding up and slowing down at different parts of the terrain. We build an offline library of periodic gaits which span two trotting steps on the robot, and switch between different motion primitives to achieve aperiodic motions of different step lengths on an A1 robot. The motion library is used to provide targets to a geometric model predictive controller which controls stance. To incorporate visual feedback, we use terrain mapping tools to build a local height map of the terrain around the robot using RGB and depth cameras, and extract feasible foothold locations around both the front and hind legs of the robot. Our experiments show a Unitree A1 robot navigating multiple unknown, challenging and discrete terrains in the real world.

Jingyun Yang, Junwu Zhang, Connor Settle et al.

We develop a method for learning periodic tasks from visual demonstrations. The core idea is to leverage periodicity in the policy structure to model periodic aspects of the tasks. We use active learning to optimize parameters of rhythmic dynamic movement primitives (rDMPs) and propose an objective to maximize the similarity between the motion of objects manipulated by the robot and the desired motion in human video demonstrations. We consider tasks with deformable objects and granular matter whose states are challenging to represent and track: wiping surfaces with a cloth, winding cables/wires, and stirring granular matter with a spoon. Our method does not require tracking markers or manual annotations. The initial training data consists of 10-minute videos of random unpaired interactions with objects by the robot and human. We use these for unsupervised learning of a keypoint model to get task-agnostic visual correspondences. Then, we use Bayesian optimization to optimize rDMPs from a single human video demonstration within few robot trials. We present simulation and hardware experiments to validate our approach.

Tianyu Li, Jungdam Won, Sehoon Ha et al.

Robots operating in human environments need various skills, like slow and fast walking, turning, side-stepping, and many more. However, building robot controllers that can exhibit such a large range of behaviors is a challenging problem that requires tedious investigation for every task. We present a unified model-based control algorithm for imitating different animal gaits without expensive simulation training or real-world fine-tuning. Our method consists of stance and swing leg controllers using a centroidal dynamics model augmented with online adaptation techniques. We also develop a whole-body trajectory optimization procedure to fix the kinematic infeasibility of the reference animal motions. We demonstrate that our universal data-driven model-based controller can seamlessly imitate various motor skills, including trotting, pacing, turning, and side-stepping. It also shows better tracking capabilities in simulation and the real world against several baselines, including another model-based imitation controller and a learning-based motion imitation technique.

Jennifer C. Shih, Akshara Rai, Laurent El Ghaoui

Multi-vehicle collision avoidance is a highly crucial problem due to the soaring interests of introducing autonomous vehicles into the real world in recent years. The safety of these vehicles while they complete their objectives is of paramount importance. Hamilton-Jacobi (HJ) reachability is a promising tool for guaranteeing safety for low-dimensional systems. However, due to its exponential complexity in computation time, no reachability-based methods have been able to guarantee safety for more than three vehicles successfully in unstructured scenarios. For systems with four or more vehicles,we can only empirically validate their safety performance.While reachability-based safety methods enjoy a flexible least-restrictive control strategy, it is challenging to reason about long-horizon trajectories online because safety at any given state is determined by looking up its safety value in a pre-computed table that does not exhibit favorable properties that continuous functions have. This motivates the problem of improving the safety performance of unstructured multi-vehicle systems when safety cannot be guaranteed given any least-restrictive safety-aware collision avoidance algorithm while avoiding online trajectory optimization. In this paper, we propose a novel approach using supervised learning to enhance the safety of vehicles by proposing new initial states in very close neighborhood of the original initial states of vehicles. Our experiments demonstrate the effectiveness of our proposed approach and show that vehicles are able to get to their goals with better safety performance with our approach compared to a baseline approach in wide-ranging scenarios.

Yixin Lin, Austin S. Wang, Eric Undersander et al.

Manipulation tasks, like loading a dishwasher, can be seen as a sequence of spatial constraints and relationships between different objects. We aim to discover these rules from demonstrations by posing manipulation as a classification problem over a graph, whose nodes represent task-relevant entities like objects and goals, and present a graph neural network (GNN) policy architecture for solving this problem from demonstrations. In our experiments, a single GNN policy trained using imitation learning (IL) on 20 expert demos can solve blockstacking, rearrangement, and dishwasher loading tasks; once the policy has learned the spatial structure, it can generalize to a larger number of objects, goal configurations, and from simulation to the real world. These experiments show that graphical IL can solve complex long-horizon manipulation problems without requiring detailed task descriptions. Videos can be found at: https://youtu.be/POxaTDAj7aY.

Joanne Truong, Denis Yarats, Tianyu Li et al.

Recent work has shown results on learning navigation policies for idealized cylinder agents in simulation and transferring them to real wheeled robots. Deploying such navigation policies on legged robots can be challenging due to their complex dynamics, and the large dynamical difference between cylinder agents and legged systems. In this work, we learn hierarchical navigation policies that account for the low-level dynamics of legged robots, such as maximum speed, slipping, contacts, and learn to successfully navigate cluttered indoor environments. To enable transfer of policies learned in simulation to new legged robots and hardware, we learn dynamics-aware navigation policies across multiple robots with robot-specific embeddings. The learned embedding is optimized on new robots, while the rest of the policy is kept fixed, allowing for quick adaptation. We train our policies across three legged robots in simulation - 2 quadrupeds (A1, AlienGo) and a hexapod (Daisy). At test time, we study the performance of our learned policy on two new legged robots in simulation (Laikago, 4-legged Daisy), and one real-world quadrupedal robot (A1). Our experiments show that our learned policy can sample-efficiently generalize to previously unseen robots, and enable sim-to-real transfer of navigation policies for legged robots.

Sarah Bechtle, Bilal Hammoud, Akshara Rai et al.

Learning for model based control can be sample-efficient and generalize well, however successfully learning models and controllers that represent the problem at hand can be challenging for complex tasks. Using inaccurate models for learning can lead to sub-optimal solutions, that are unlikely to perform well in practice. In this work, we present a learning approach which iterates between model learning and data collection and leverages forward model prediction error for learning control. We show how using the controller's prediction as input to a forward model can create a differentiable connection between the controller and the model, allowing us to formulate a loss in the state space. This lets us include forward model prediction error during controller learning and we show that this creates a loss objective that significantly improves learning on different motor control tasks. We provide empirical and theoretical results that show the benefits of our method and present evaluations in simulation for learning control on a 7 DoF manipulator and an underactuated 12 DoF quadruped. We show that our approach successfully learns controllers for challenging motor control tasks involving contact switching.

Neha Das, Sarah Bechtle, Todor Davchev et al.

Scaling model-based inverse reinforcement learning (IRL) to real robotic manipulation tasks with unknown dynamics remains an open problem. The key challenges lie in learning good dynamics models, developing algorithms that scale to high-dimensional state-spaces and being able to learn from both visual and proprioceptive demonstrations. In this work, we present a gradient-based inverse reinforcement learning framework that utilizes a pre-trained visual dynamics model to learn cost functions when given only visual human demonstrations. The learned cost functions are then used to reproduce the demonstrated behavior via visual model predictive control. We evaluate our framework on hardware on two basic object manipulation tasks.

Tianyu Li, Roberto Calandra, Deepak Pathak et al.

Hierarchical learning has been successful at learning generalizable locomotion skills on walking robots in a sample-efficient manner. However, the low-dimensional "latent" action used to communicate between two layers of the hierarchy is typically user-designed. In this work, we present a fully-learned hierarchical framework, that is capable of jointly learning the low-level controller and the high-level latent action space. Once this latent space is learned, we plan over continuous latent actions in a model-predictive control fashion, using a learned high-level dynamics model. This framework generalizes to multiple robots, and we present results on a Daisy hexapod simulation, A1 quadruped simulation, and Daisy robot hardware. We compare a range of learned hierarchical approaches from literature, and show that our framework outperforms baselines on multiple tasks and two simulations. In addition to learning approaches, we also compare to inverse-kinematics (IK) acting on desired robot motion, and show that our fully-learned framework outperforms IK in adverse settings on both A1 and Daisy simulations. On hardware, we show the Daisy hexapod achieve multiple locomotion tasks, in an unstructured outdoor setting, with only 2000 hardware samples, reinforcing the robustness and sample-efficiency of our approach.

Kristen Morse, Neha Das, Yixin Lin et al.

Being able to quickly adapt to changes in dynamics is paramount in model-based control for object manipulation tasks. In order to influence fast adaptation of the inverse dynamics model's parameters, data efficiency is crucial. Given observed data, a key element to how an optimizer updates model parameters is the loss function. In this work, we propose to apply meta-learning to learn structured, state-dependent loss functions during a meta-training phase. We then replace standard losses with our learned losses during online adaptation tasks. We evaluate our proposed approach on inverse dynamics learning tasks, both in simulation and on real hardware data. In both settings, the structured and state-dependent learned losses improve online adaptation speed, when compared to standard, state-independent loss functions.

Benjamin Letham, Roberto Calandra, Akshara Rai et al.

Bayesian optimization (BO) is a popular approach to optimize expensive-to-evaluate black-box functions. A significant challenge in BO is to scale to high-dimensional parameter spaces while retaining sample efficiency. A solution considered in existing literature is to embed the high-dimensional space in a lower-dimensional manifold, often via a random linear embedding. In this paper, we identify several crucial issues and misconceptions about the use of linear embeddings for BO. We study the properties of linear embeddings from the literature and show that some of the design choices in current approaches adversely impact their performance. We show empirically that properly addressing these issues significantly improves the efficacy of linear embeddings for BO on a range of problems, including learning a gait policy for robot locomotion.

Giovanni Sutanto, Austin S. Wang, Yixin Lin et al.

The recursive Newton-Euler Algorithm (RNEA) is a popular technique for computing the dynamics of robots. RNEA can be framed as a differentiable computational graph, enabling the dynamics parameters of the robot to be learned from data via modern auto-differentiation toolboxes. However, the dynamics parameters learned in this manner can be physically implausible. In this work, we incorporate physical constraints in the learning by adding structure to the learned parameters. This results in a framework that can learn physically plausible dynamics via gradient descent, improving the training speed as well as generalization of the learned dynamics models. We evaluate our method on real-time inverse dynamics control tasks on a 7 degree of freedom robot arm, both in simulation and on the real robot. Our experiments study a spectrum of structure added to the parameters of the differentiable RNEA algorithm, and compare their performance and generalization.

Tianyu Li, Nathan Lambert, Roberto Calandra et al.

Learning to locomote to arbitrary goals on hardware remains a challenging problem for reinforcement learning. In this paper, we present a hierarchical learning framework that improves sample-efficiency and generalizability of locomotion skills on real-world robots. Our approach divides the problem of goal-oriented locomotion into two sub-problems: learning diverse primitives skills, and using model-based planning to sequence these skills. We parametrize our primitives as cyclic movements, improving sample-efficiency of learning on a 18 degrees of freedom robot. Then, we learn coarse dynamics models over primitive cycles and use them in a model predictive control framework. This allows us to learn to walk to arbitrary goals up to 12m away, after about two hours of training from scratch on hardware. Our results on a Daisy hexapod hardware and simulation demonstrate the efficacy of our approach at reaching distant targets, in different environments and with sensory noise.

Rika Antonova, Akshara Rai, Tianyu Li et al.

Data-efficiency is crucial for autonomous robots to adapt to new tasks and environments. In this work we focus on robotics problems with a budget of only 10-20 trials. This is a very challenging setting even for data-efficient approaches like Bayesian optimization (BO), especially when optimizing higher-dimensional controllers. Simulated trajectories can be used to construct informed kernels for BO. However, previous work employed supervised ways of extracting low-dimensional features for these. We propose a model and architecture for a sequential variational autoencoder that embeds the space of simulated trajectories into a lower-dimensional space of latent paths in an unsupervised way. We further compress the search space for BO by reducing exploration in parts of the state space that are undesirable, without requiring explicit constraints on controller parameters. We validate our approach with hardware experiments on a Daisy hexapod robot and an ABB Yumi manipulator. We also present simulation experiments with further comparisons to several baselines on Daisy and two manipulators. Our experiments indicate the proposed trajectory-based kernel with dynamic compression can offer ultra data-efficient optimization.

Sarah Bechtle, Yixin Lin, Akshara Rai et al.

Curiosity as a means to explore during reinforcement learning problems has recently become very popular. However, very little progress has been made in utilizing curiosity for learning control. In this work, we propose a model-based reinforcement learning (MBRL) framework that combines Bayesian modeling of the system dynamics with curious iLQR, an iterative LQR approach that considers model uncertainty. During trajectory optimization the curious iLQR attempts to minimize both the task-dependent cost and the uncertainty in the dynamics model. We demonstrate the approach on reaching tasks with 7-DoF manipulators in simulation and on a real robot. Our experiments show that MBRL with curious iLQR reaches desired end-effector targets more reliably and with less system rollouts when learning a new task from scratch, and that the learned model generalizes better to new reaching tasks.

Tianyu Li, Akshara Rai, Hartmut Geyer et al.

Learning controllers for bipedal robots is a challenging problem, often requiring expert knowledge and extensive tuning of parameters that vary in different situations. Recently, deep reinforcement learning has shown promise at automatically learning controllers for complex systems in simulation. This has been followed by a push towards learning controllers that can be transferred between simulation and hardware, primarily with the use of domain randomization. However, domain randomization can make the problem of finding stable controllers even more challenging, especially for underactuated bipedal robots. In this work, we explore whether policies learned in simulation can be transferred to hardware with the use of high-fidelity simulators and structured controllers. We learn a neural network policy which is a part of a more structured controller. While the neural network is learned in simulation, the rest of the controller stays fixed, and can be tuned by the expert as needed. We show that using this approach can greatly speed up the rate of learning in simulation, as well as enable transfer of policies between simulation and hardware. We present our results on an ATRIAS robot and explore the effect of action spaces and cost functions on the rate of transfer between simulation and hardware. Our results show that structured policies can indeed be learned in simulation and implemented on hardware successfully. This has several advantages, as the structure preserves the intuitive nature of the policy, and the neural network improves the performance of the hand-designed policy. In this way, we propose a way of using neural networks to improve expert designed controllers, while maintaining ease of understanding.

Akshara Rai, Rika Antonova, Franziska Meier et al.

Learning for control can acquire controllers for novel robotic tasks, paving the path for autonomous agents. Such controllers can be expert-designed policies, which typically require tuning of parameters for each task scenario. In this context, Bayesian optimization (BO) has emerged as a promising approach for automatically tuning controllers. However, when performing BO on hardware for high-dimensional policies, sample-efficiency can be an issue. Here, we develop an approach that utilizes simulation to map the original parameter space into a domain-informed space. During BO, similarity between controllers is now calculated in this transformed space. Experiments on the ATRIAS robot hardware and another bipedal robot simulation show that our approach succeeds at sample-efficiently learning controllers for multiple robots. Another question arises: What if the simulation significantly differs from hardware? To answer this, we create increasingly approximate simulators and study the effect of increasing simulation-hardware mismatch on the performance of Bayesian optimization. We also compare our approach to other approaches from literature, and find it to be more reliable, especially in cases of high mismatch. Our experiments show that our approach succeeds across different controller types, bipedal robot models and simulator fidelity levels, making it applicable to a wide range of bipedal locomotion problems.

Akshara Rai, Rika Antonova, Seungmoon Song et al.

Controllers in robotics often consist of expert-designed heuristics, which can be hard to tune in higher dimensions. It is typical to use simulation to learn these parameters, but controllers learned in simulation often don't transfer to hardware. This necessitates optimization directly on hardware. However, collecting data on hardware can be expensive. This has led to a recent interest in adapting data-efficient learning techniques to robotics. One popular method is Bayesian Optimization (BO), a sample-efficient black-box optimization scheme, but its performance typically degrades in higher dimensions. We aim to overcome this problem by incorporating domain knowledge to reduce dimensionality in a meaningful way, with a focus on bipedal locomotion. In previous work, we proposed a transformation based on knowledge of human walking that projected a 16-dimensional controller to a 1-dimensional space. In simulation, this showed enhanced sample efficiency when optimizing human-inspired neuromuscular walking controllers on a humanoid model. In this paper, we present a generalized feature transform applicable to non-humanoid robot morphologies and evaluate it on the ATRIAS bipedal robot -- in simulation and on hardware. We present three different walking controllers; two are evaluated on the real robot. Our results show that this feature transform captures important aspects of walking and accelerates learning on hardware and simulation, as compared to traditional BO.

Rika Antonova, Akshara Rai, Christopher G. Atkeson

Sample efficiency is important when optimizing parameters of locomotion controllers, since hardware experiments are time consuming and expensive. Bayesian Optimization, a sample-efficient optimization framework, has recently been widely applied to address this problem, but further improvements in sample efficiency are needed for practical applicability to real-world robots and high-dimensional controllers. To address this, prior work has proposed using domain expertise for constructing custom distance metrics for locomotion. In this work we show how to learn such a distance metric automatically. We use a neural network to learn an informed distance metric from data obtained in high-fidelity simulations. We conduct experiments on two different controllers and robot architectures. First, we demonstrate improvement in sample efficiency when optimizing a 5-dimensional controller on the ATRIAS robot hardware. We then conduct simulation experiments to optimize a 16-dimensional controller for a 7-link robot model and obtain significant improvements even when optimizing in perturbed environments. This demonstrates that our approach is able to enhance sample efficiency for two different controllers, hence is a fitting candidate for further experiments on hardware in the future.

Rika Antonova, Akshara Rai, Christopher G. Atkeson

Learning policies for bipedal locomotion can be difficult, as experiments are expensive and simulation does not usually transfer well to hardware. To counter this, we need al- gorithms that are sample efficient and inherently safe. Bayesian Optimization is a powerful sample-efficient tool for optimizing non-convex black-box functions. However, its performance can degrade in higher dimensions. We develop a distance metric for bipedal locomotion that enhances the sample-efficiency of Bayesian Optimization and use it to train a 16 dimensional neuromuscular model for planar walking. This distance metric reflects some basic gait features of healthy walking and helps us quickly eliminate a majority of unstable controllers. With our approach we can learn policies for walking in less than 100 trials for a range of challenging settings. In simulation, we show results on two different costs and on various terrains including rough ground and ramps, sloping upwards and downwards. We also perturb our models with unknown inertial disturbances analogous with differences between simulation and hardware. These results are promising, as they indicate that this method can potentially be used to learn control policies on hardware.

Akshara Rai, Giovanni Sutanto, Stefan Schaal et al.

With the advancement of robotics, machine learning, and machine perception, increasingly more robots will enter human environments to assist with daily tasks. However, dynamically-changing human environments requires reactive motion plans. Reactivity can be accomplished through replanning, e.g. model-predictive control, or through a reactive feedback policy that modifies on-going behavior in response to sensory events. In this paper, we investigate how to use machine learning to add reactivity to a previously learned nominal skilled behavior. We approach this by learning a reactive modification term for movement plans represented by nonlinear differential equations. In particular, we use dynamic movement primitives (DMPs) to represent a skill and a neural network to learn a reactive policy from human demonstrations. We use the well explored domain of obstacle avoidance for robot manipulation as a test bed. Our approach demonstrates how a neural network can be combined with physical insights to ensure robust behavior across different obstacle settings and movement durations. Evaluations on an anthropomorphic robotic system demonstrate the effectiveness of our work.