Robot-assisted dressing offers an opportunity to benefit the lives of many people with disabilities, such as some older adults. However, robots currently lack common sense about the physical implications of their actions on people. The physical implications of dressing are complicated by non-rigid garments, which can result in a robot indirectly applying high forces to a person's body. We present a deep recurrent model that, when given a proposed action by the robot, predicts the forces a garment will apply to a person's body. We also show that a robot can provide better dressing assistance by using this model with model predictive control. The predictions made by our model only use haptic and kinematic observations from the robot's end effector, which are readily attainable. Collecting training data from real world physical human-robot interaction can be time consuming, costly, and put people at risk. Instead, we train our predictive model using data collected in an entirely self-supervised fashion from a physics-based simulation. We evaluated our approach with a PR2 robot that attempted to pull a hospital gown onto the arms of 10 human participants. With a 0.2s prediction horizon, our controller succeeded at high rates and lowered applied force while navigating the garment around a persons fist and elbow without getting caught. Shorter prediction horizons resulted in significantly reduced performance with the sleeve catching on the participants' fists and elbows, demonstrating the value of our model's predictions. These behaviors of mitigating catches emerged from our deep predictive model and the controller objective function, which primarily penalizes high forces.

No abstract

Robots could be a valuable tool for helping with dressing but determining how a robot and a person with disabilities can collaborate to complete the task is challenging. We present task optimization of robot-assisted dressing (TOORAD), a method for generating a plan that consists of actions for both the robot and the person. TOORAD uses a multilevel optimization framework with heterogeneous simulations. The simulations model the physical interactions between the garment and the person being dressed, as well as the geometry and kinematics of the robot, human, and environment. Notably, the models for the human are personalized for an individual's geometry and physical capabilities. TOORAD searches over a constrained action space that interleaves the motions of the person and the robot with the person remaining still when the robot moves and vice versa. In order to adapt to real-world variation, TOORAD incorporates a measure of robot dexterity in its optimization, and the robot senses the person's body with a capacitive sensor to adapt its planned end effector trajectories. To evaluate TOORAD and gain insight into robot-assisted dressing, we conducted a study with six participants with physical disabilities who have difficulty dressing themselves. In the first session, we created models of the participants and surveyed their needs, capabilities, and views on robot-assisted dressing. TOORAD then found personalized plans and generated instructional visualizations for four of the participants, who returned for a second session during which they successfully put on both sleeves of a hospital gown with assistance from the robot. Overall, our work demonstrates the feasibility of generating personalized plans for robot-assisted dressing via optimization and physics-based simulation.

Robotic assistance presents an opportunity to benefit the lives of many people with physical disabilities, yet accurately sensing the human body and tracking human motion remain difficult for robots. We present a multidimensional capacitive sensing technique that estimates the local pose of a human limb in real time. A key benefit of this sensing method is that it can sense the limb through opaque materials, including fabrics and wet cloth. Our method uses a multielectrode capacitive sensor mounted to a robot's end effector. A neural network model estimates the position of the closest point on a person's limb and the orientation of the limb's central axis relative to the sensor's frame of reference. These pose estimates enable the robot to move its end effector with respect to the limb using feedback control. We demonstrate that a PR2 robot can use this approach with a custom six electrode capacitive sensor to assist with two activities of daily livingdressing and bathing. The robot pulled the sleeve of a hospital gown onto able-bodied participants' right arms, while tracking human motion. When assisting with bathing, the robot moved a soft wet washcloth to follow the contours of able-bodied participants' limbs, cleaning their surfaces. Overall, we found that multidimensional capacitive sensing presents a promising approach for robots to sense and track the human body during assistive tasks that require physical human-robot interaction.

No abstract