12 May 2021 in Science Robotics
A simple robotic hand leverages sliding, rolling, and contact breaking of objects and caging for highly dexterous manipulation.
12 May 2021 in Science Robotics
Self-identification enables robotic manipulation systems to interactively acquire model parameters with limited sensing.
19 December 2019 in IEEE Robotics and Automation Letters
In this work, we propose a pick-and-place benchmark to assess the manipulation capabilities of a robotic system. The benchmark is based on the Box and Blocks Test (BBT), a task utilized for decades by the rehabilitation community to assess unilateral gross manual dexterity in humans. We propose three robot benchmarking protocols in this work that hold true to the spirit of the original clinical tests-the Modified-BBT, the Targeted-BBT, and the Standard-BBT. These protocols can be implemented by the greater robotics research community, as the physical BBT setup has been widely distributed with the Yale-CMU-Berkeley (YCB) Object and Model Set. Difficulty of the three protocols increase sequentially, adding a new performance component at each level, and therefore aiming to assess various aspects of the system separately. Clinical task-time norms are summarized for able-bodied human participants. We provide baselines for all three protocols with off-the-shelf planning and perception algorithms on a Barrett WAM and a Franka Emika Panda manipulator, and compare results with human performance.
25 January 2019 in IEEE Transactions on Robotics
Human in-hand dexterity can be highly fluid and unstructured, with multiple phalanxes breaking and re-establishing contact during any given task. In contrast, prevailing research in robotic manipulation has focused on highly structured well-controlled motions, where contact points are carefully characterized. Maintaining grasp stability by satisfying traditional closure conditions during complex within-hand manipulation motions can be difficult, even with highly articulated end effectors. However, simple grippers can still achieve an effective range of in-hand manipulation tasks without strict closure conditions, as long as the object can be bounded locally relative to the hand frame. The end effector can be considered as a tool to limit the range of possible object poses. In particular, the energy of the hand-object system can be used to determine an attractor region toward which the hand drives the object. This can be combined with a sparse sampling of the configuration space to find a set of manipulation primitives that can reliably constrain the object inside the hand workspace even without feedback, a strategy proposed as whole-hand caging manipulation. In this paper, experimental results with a planar underactuated gripper are presented to validate this manipulation strategy, and it is shown that even though contacts are regularly broken and reformed, the object can be reliably manipulated within the hand workspace without ejection, and challenging movements such as sliding and gaiting can be reliably performed.
11 March 2021 in IEEE Transactions on Robotics
Learning accurate representations of robot models remains a challenging problem, and is typically approached though large, system-specific feature sets. This method inherently introduces practical shortcomings, as interpretability and transferability of the learned model typically decreases as more features are introduced into the learning framework in order to handle increasing task complexity. In this article, we examine the problem of developing transferable learned models for dexterous manipulation that are able to accurately predict the behavior of physically distinct systems without retraining. We introduce the notion of learning from visually-extracted grasp mechanics-based features, which are formulated by combining geometrically-inspired, analytical representations of the gripper into the feature set to more holistically represent the state of varied systems performing manipulation. We characterize the added utility of using such features through simulation and incorporate them into a classifier to predict specific phenomena, or modes of manipulation, that occur during prehensile within-hand movement. Four modes of manipulation—normal (rolling contact), drop, stuck, and sliding—are defined, collected physically, and trained via a self-supervised learning approach. The classifier is first trained on a single sensorless underactuated hand variant for all four modes. We, then, investigate the transferability of the learned classifier on five different planar gripper variants—analyzing applicability of this approach with both online and offline evaluation.