Incremental Learning for Robot Shared Autonomy

Shared autonomy holds promise for improving the usability and accessibility of assistive robotic arms, but current methods often rely on costly expert demonstrations and lack the ability to adapt post-deployment. This paper introduces ILSA, an Incrementally Learned Shared Autonomy framework that continually improves its assistive control policy through repeated user interactions. ILSA leverages synthetic kinematic trajectories for initial pretraining, reducing the need for expert demonstrations, and then incrementally finetunes its policy after each manipulation interaction, with mechanisms to balance new knowledge acquisition with existing knowledge retention during incremental learning. We validate ILSA for complex long-horizon tasks through a comprehensive ablation study and a user study with 20 participants, demonstrating its effectiveness and robustness in both quantitative performance and user-reported qualitative metrics. Code and videos are available at https://ilsa-robo.github.io/.

Do Mistakes Matter? Comparing Trust Responses of Different Age Groups to Errors Made by Physically Assistive Robots

Trust is a key factor in ensuring acceptable human-robot interaction, especially in settings where robots may be assisting with critical activities of daily living. When practically deployed, robots are bound to make occasional mistakes, yet the degree to which these errors will impact a care recipient's trust in the robot, especially in performing physically assistive tasks, remains an open question. To investigate this, we conducted experiments where participants interacted with physically assistive robots which would occasionally make intentional mistakes while performing two different tasks: bathing and feeding. Our study considered the error response of two populations: younger adults at a university (median age 26) and older adults at an independent living facility (median age 83). We observed that the impact of errors on a users' trust in the robot depends on both their age and the task that the robot is performing. We also found that older adults tend to evaluate the robot on factors unrelated to the robot's performance, making their trust in the system more resilient to errors when compared to younger adults. Code and supplementary materials are available on our project webpage.

Unfolding the Literature: A Review of Robotic Cloth Manipulation

The realm of textiles spans clothing, households, healthcare, sports, and industrial applications. The deformable nature of these objects poses unique challenges that prior work on rigid objects cannot fully address. The increasing interest within the community in textile perception and manipulation has led to new methods that aim to address challenges in modeling, perception, and control, resulting in significant progress. However, this progress is often tailored to one specific textile or a subcategory of these textiles. To understand what restricts these methods and hinders current approaches from generalizing to a broader range of real-world textiles, this review provides an overview of the field, focusing specifically on how and to what extent textile variations are addressed in modeling, perception, benchmarking, and manipulation of textiles. We finally conclude by identifying key open problems and outlining grand challenges that will drive future advancements in the field.

RoboCAP: Robotic Classification and Precision Pouring of Diverse Liquids and Granular Media with Capacitive Sensing

Liquids and granular media are pervasive throughout human environments, yet remain particularly challenging for robots to sense and manipulate precisely. In this work, we present a systematic approach at integrating capacitive sensing within robotic end effectors to enable robust sensing and precise manipulation of liquids and granular media. We introduce the parallel-jaw RoboCAP Gripper with embedded capacitive sensing arrays that enable a robot to directly sense the materials and dynamics of liquids inside of diverse containers, including some visually opaque. When coupled with model-based control, we demonstrate that the proposed system enables a robotic manipulator to achieve state-of-the-art precision pouring accuracy for a range of substances with varying dynamics properties. Code, designs, and build details are available on the project website.

SkinGrip: An Adaptive Soft Robotic Manipulator with Capacitive Sensing for Whole-Limb Bathing Assistance

Robotics presents a promising opportunity for enhancing bathing assistance, potentially to alleviate labor shortages and reduce care costs, while offering consistent and gentle care for individuals with physical disabilities. However, ensuring flexible and efficient cleaning of the human body poses challenges as it involves direct physical contact between the human and the robot, and necessitates simple, safe, and effective control. In this paper, we introduce a soft, expandable robotic manipulator with embedded capacitive proximity sensing arrays, designed for safe and efficient bathing assistance. We conduct a thorough evaluation of our soft manipulator, comparing it with a baseline rigid end effector in a human study involving 12 participants across $96$ bathing trails. Our soft manipulator achieves an an average cleaning effectiveness of 88.8% on arms and 81.4% on legs, far exceeding the performance of the baseline. Participant feedback further validates the manipulator's ability to maintain safety, comfort, and thorough cleaning.

VoicePilot: Harnessing LLMs as Speech Interfaces for Physically Assistive Robots

Physically assistive robots present an opportunity to significantly increase the well-being and independence of individuals with motor impairments or other forms of disability who are unable to complete activities of daily living. Speech interfaces, especially ones that utilize Large Language Models (LLMs), can enable individuals to effectively and naturally communicate high-level commands and nuanced preferences to robots. Frameworks for integrating LLMs as interfaces to robots for high level task planning and code generation have been proposed, but fail to incorporate human-centric considerations which are essential while developing assistive interfaces. In this work, we present a framework for incorporating LLMs as speech interfaces for physically assistive robots, constructed iteratively with 3 stages of testing involving a feeding robot, culminating in an evaluation with 11 older adults at an independent living facility. We use both quantitative and qualitative data from the final study to validate our framework and additionally provide design guidelines for using LLMs as speech interfaces for assistive robots. Videos and supporting files are located on our project website: https://sites.google.com/andrew.cmu.edu/voicepilot/

BodyMAP -- Jointly Predicting Body Mesh and 3D Applied Pressure Map for People in Bed

Accurately predicting the 3D human posture and the pressure exerted on the body for people resting in bed, visualized as a body mesh (3D pose & shape) with a 3D pressure map, holds significant promise for healthcare applications, particularly, in the prevention of pressure ulcers. Current methods focus on singular facets of the problem -- predicting only 2D/3D poses, generating 2D pressure images, predicting pressure only for certain body regions instead of the full body, or forming indirect approximations to the 3D pressure map. In contrast, we introduce BodyMAP, which jointly predicts the human body mesh and 3D applied pressure map across the entire human body. Our network leverages multiple visual modalities, incorporating both a depth image of a person in bed and its corresponding 2D pressure image acquired from a pressure-sensing mattress. The 3D pressure map is represented as a pressure value at each mesh vertex and thus allows for precise localization of high-pressure regions on the body. Additionally, we present BodyMAP-WS, a new formulation of pressure prediction in which we implicitly learn pressure in 3D by aligning sensed 2D pressure images with a differentiable 2D projection of the predicted 3D pressure maps. In evaluations with real-world human data, our method outperforms the current state-of-the-art technique by 25% on both body mesh and 3D applied pressure map prediction tasks for people in bed.

AdaFold: Adapting Folding Trajectories of Cloths via Feedback-loop Manipulation

We present AdaFold, a model-based feedback-loop framework for optimizing folding trajectories. AdaFold extracts a particle-based representation of cloth from RGB-D images and feeds back the representation to a model predictive control to re-plan folding trajectory at every time-step. A key component of AdaFold that enables feedback-loop manipulation is the use of semantic descriptors extracted from visual-language models. These descriptors enhance the particle representation of the cloth to distinguish between ambiguous point clouds of differently folded cloths. Our experiments demonstrate AdaFold's ability to adapt folding trajectories to cloths with varying physical properties and generalize from simulated training to real-world execution.

RL-VLM-F: Reinforcement Learning from Vision Language Foundation Model Feedback

Reward engineering has long been a challenge in Reinforcement Learning (RL) research, as it often requires extensive human effort and iterative processes of trial-and-error to design effective reward functions. In this paper, we propose RL-VLM-F, a method that automatically generates reward functions for agents to learn new tasks, using only a text description of the task goal and the agent's visual observations, by leveraging feedbacks from vision language foundation models (VLMs). The key to our approach is to query these models to give preferences over pairs of the agent's image observations based on the text description of the task goal, and then learn a reward function from the preference labels, rather than directly prompting these models to output a raw reward score, which can be noisy and inconsistent. We demonstrate that RL-VLM-F successfully produces effective rewards and policies across various domains - including classic control, as well as manipulation of rigid, articulated, and deformable objects - without the need for human supervision, outperforming prior methods that use large pretrained models for reward generation under the same assumptions.

DiffTOP: Differentiable Trajectory Optimization for Deep Reinforcement and Imitation Learning

This paper introduces DiffTOP, which utilizes Differentiable Trajectory OPtimization as the policy representation to generate actions for deep reinforcement and imitation learning. Trajectory optimization is a powerful and widely used algorithm in control, parameterized by a cost and a dynamics function. The key to our approach is to leverage the recent progress in differentiable trajectory optimization, which enables computing the gradients of the loss with respect to the parameters of trajectory optimization. As a result, the cost and dynamics functions of trajectory optimization can be learned end-to-end. DiffTOP addresses the ``objective mismatch'' issue of prior model-based RL algorithms, as the dynamics model in DiffTOP is learned to directly maximize task performance by differentiating the policy gradient loss through the trajectory optimization process. We further benchmark DiffTOP for imitation learning on standard robotic manipulation task suites with high-dimensional sensory observations and compare our method to feed-forward policy classes as well as Energy-Based Models (EBM) and Diffusion. Across 15 model-based RL tasks and 13 imitation learning tasks with high-dimensional image and point cloud inputs, DiffTOP outperforms prior state-of-the-art methods in both domains.