Cloth physical properties and speed drive variability in dynamic cloth motions
Same Action, Different Outcome: Variability in Dynamic Cloth Manipulation
Robotics
Summary
Cloth behaves unpredictably when moved quickly along the same path, and this paper studies how much the outcomes can differ. The authors measured how much cloth movement varies by repeating tasks many times with different cloth types and speeds. They found that the cloth's material and the speed of motion mainly cause this variability, which is much larger than robot or sensor inconsistencies. They also tested digital cloth simulators and found none can fully capture this real-world variability. The authors provide detailed data to help improve future cloth manipulation and simulation.
What this means in practice
- •For robotics engineers: Improve robot manipulation policies by accounting for cloth variability in physical properties and motion speeds when designing control algorithms.
- •For simulation developers: Develop better cloth simulators by using the published dataset to calibrate models that realistically reproduce real-world variability in cloth behavior.
Authors
Mahed Dadgostar, Guillem Alenyà, Júlia Borràs
Abstract
Although cloth is known to exhibit different outcomes under repeated fast dynamic motions, even when the same trajectory is applied, this variability has not yet been systematically characterized. Quantifying it is essential to assess the reliability of learned manipulation policies and the extent to which simulation can reproduce real-world behavior. To study this, we execute the same trajectory ten times across four dynamic tasks, two of which are novel, each tested with three cloths of very different properties and at up to three execution speeds, with a total of 269 recorded rollouts. For all of them, we record small marker positions on the cloth and synchronized stereo camera. We then formalize different metrics to quantify variability, and our results show how it is significant in every test condition, in most cases one to three orders of magnitude above the repeatability inherent to the robot and sensing noise. Our results also show variability is driven mainly by the cloth physical properties but also grows with speed. By replaying all the rollouts in four calibrated modern cloth simulators, we show that none of them can reproduce the variability magnitude we observed in real cloth, nor its ordering. Together with our analysis, we publish the dataset with synchronized OptiTrack, vision and robot logs and their corresponding simulator twins.