Dexterous Manipulation of Cloth

Abstract

Figure 1: Given a simple description of the desired cloth motion, our algorithm computes appropriate joint torques for a physically simulated hand, such that, via contact forces, the result of cloth simulation follows the desired motion. AbstractThis paper introduces a new technique to synthesize dexterous manipulation of cloth. Given a simple description of the desired cloth motion, our algorithm computes appropriate joint torques for physically simulated hands, such that, via contact forces, the result of cloth simulation follows the desired motion. Instead of optimizing the hand control forces directly, we formulate an optimization problem that solves for the commanding forces from the hands to the cloth, which have more direct impact on the dynamic state of the hands and that of the cloth. The solution of the optimization provides commanding forces that achieve the desired cloth motion described by the user, while respecting the kinematic constraints of the hands. These commanding forces are then used to guide the joint torques of the hands. To balance between the effectiveness of control and computational costs, we formulate a model-predictive-control problem as a quadratic program at each time step. We demonstrate our technique on a set of cloth manipulation tasks in daily activities, including folding laundry, wringing a towel, and putting on a scarf.

An implicit frictional contact solver for adaptive cloth simulation

“…In the Shawl animation (Figure 13), we show that our friction solver can be used to model manual manipulation of cloth, in this case, throwing a shawl over one shoulder. Previous work on dexterous manipulation [Bai et al 2016;Clegg et al 2015] has required artificial sticking models or additional constraints due to inadequacies with existing contact models for cloth. In this animation, we manually animated only the motion of the character's arm and hand, while the shawl was simulated using our method.…”

“…For instance, beyond what achieved today [Sigal et al 2015], an artist should be able to tune a friction coefficient finely so as to explore wider ranges of cloth materials, and accurately experience stick-slip transitions during motion. Likewise, an avatar should be able to grasp a garment thanks to Coulomb friction only, without having to resort to some artificial sticking model at pinch [Clegg et al 2015] nor to the enforcement of static friction at controlled vertices [Bai et al 2016]. Applications at the frontier with nearby disciplines like engineering, robotics or medicine also call for much more accurate models for frictional contact than what are currently used.…”

Multi-Modal Sensing and Robotic Manipulation of Non-Rigid Objects: A Survey

“…This was later improved in Higashimori et al [83] by allowing the object to "jump" from the plate during shaping. From the realm of computer graphics and animation, Bai et al [84] developed an algorithm to compute the necessary joint torques of simulated anthropomorphic hands in order for a cloth to follow a defined motion. It is interesting that the task description is given as a path to be followed by any point(s) of the cloth.…”

Learning When to Trust a Dynamics Model for Planning in Reduced State Spaces

“…Since the commands are the same, our rollout function Γ, which uses u b , is equivalent to f . To find a path in B we must solve Problem (2). We use the planner described in [17] to solve this problem; this is an RRT-based planner designed for use with virtual elastic bands as part of the planning configuration space.…”

“…This simulation can be time-consuming and/or inaccurate. Including such simulations inside a planner can result in plans that take hours to compute [2].…”