Iterative Learning of Grasp Adaptation through Human Corrections

In the context of object interaction and manipulation, one characteristic of a robust grasp is its ability to comply with external perturbations applied to the grasped object while still maintaining the grasp. In this work we introduce an approach for grasp adaptation which learns a statistical model to adapt hand posture solely based on the perceived contact between the object and fingers. Using a multi-step learning procedure, the model dataset is built by first demonstrating an initial hand posture, which is then physically corrected by a human teacher pressing on the fingertips, exploiting compliance in the robot hand. The learner then replays the resulting sequence of hand postures, to generate a dataset of posture-contact pairs that are not influenced by the touch of the teacher. A key feature of this work is that the learned model may be further refined by repeating the correction-replay steps. Alternatively, the model may be reused in the development of new models, characterized by the contact signatures of a different object. Our approach is empirically validated on the iCub robot. We demonstrate grasp adaptation in response to changes in contact, and show successful model reuse and improved adaptation with additional rounds of model refinement.

Control and Learning of Compliant Manipulation Skills

Klas Jonas Alfred Kronander

Humans demonstrate an impressive capability to manipulate fragile objects without damaging them, graciously controlling the force and position of hands or tools. Traditionally, robotics has favored position control over force control to produce fast, accurate and repeatable motion. For extending the applicability of robotic manipulators outside the strictly controlled environments of industrial work cells, position control is inadequate. Tasks that involve contact with objects whose positions are not known with perfect certainty require a controller that regulates the relationship between positional deviations and forces on the robot. This problem is formalized in the impedance control framework, which focuses the robot control problem on the interaction between the robot and its environment. By adjusting the impedance parameters, the behavior of the robot can be adapted to the need of the task. However, it is often difficult to specify formally how the impedance should vary for best performance. Furthermore, fast it can be shown that careless variation of the impedance can lead to unstable regulation or tracking even in free motion. In the first part of the thesis, the problem of how to define a varying impedance for a task is addressed. A haptic human-robot interface that allows a human supervisor to teach impedance variations by physically interacting with the robot during task execution is introduced. It is shown that the interface can be used to enhance the performance in several manipulation tasks. Then, the problem of stable control with varying impedance is addressed. Along with a theoretical discussion on this topic, a sufficient condition for stable varying stiffness and damping is provided. In the second part of the thesis, we explore more complex manipulation scenarios via online generation of the robot trajectory. This is done along two axes 1) learning how to react to contact forces in insertion tasks which are crucial for assembly operations and 2) autonomous Dynamical Systems (DS) for motion representation with the capability to encode a family of trajectories rather than a fixed, time-dependent reference. A novel framework for task representation using DS is introduced, termed Locally Modulated Dynamical Systems (LMDS). LMDS differs from existing DS estimation algorithms in that it supports non-parametric and incremental learning all the while guaranteeing that the resulting DS is globally stable at an attractor point. To combine the advantages of DS motion generation with impedance control, a novel controller for tasks described by first order DS is proposed. The controller is passive, and has the properties of an impedance controller with the added flexibility of a DS motion representation instead of a time-indexed trajectory.

EPFL2015

Dynamic Grasp Adaptation

Miao Li

The human hand is an amazing tool, demonstrated by its incredible motor capability and remarkable sense of touch. To enable robots to work in a human-centric environment, it is desirable to endow robotic hands with human-like capabilities for grasping and object manipulation. However, due to its inherent complexity and inevitable model uncertainty, robotic grasping and manipulation remains a challenge. This thesis focuses on grasp adaptation in the face of model and sensing uncertainties: Given an object whose properties are not known with certainty (e.g., shape, weight and external perturbation), and a multifingered robotic hand, we aim at determining where to put the fingers and how the fingers should adaptively interact with the object using tactile sensing, in order to achieve either a stable grasp or a desired dynamic behaviour. A central idea in this thesis is the object-centric dynamics: namely, that we express all control constraints into an object-centric representation. This simplifies computa- tion and makes the control versatile to the type of hands. This is an essential feature that distinguishes our work from other robust grasping work in the literature, where generating a static stable grasp for a given hand is usually the primary goal. In this thesis, grasp adaptation is a dynamic process that flexibly adapts the grasp to fit some purpose from the objectâs perspective, in the presence of a variety of uncertainties and/or perturbations. When building a grasp adaptation for a given situation, there are two key problems that must be addressed: 1) the problem of choosing an initial grasp that is suitable for future adaptation, and more importantly 2) the problem of design- ing an adaptation strategy that can react adequately to achieve desired behaviour of the grasped object. To address challenge 1 (planning a grasp under shape uncertainty), we propose an approach to parameterizing the uncertainty in object shape using Gaussian Processes (GPs) and incorporate it as a constraint into contact-level grasp planning. To realize the planned contacts using different hands interchangeably, we further develop a prob- abilistic model to predict the feasible hand configurations, including hand pose and finger joints, given the desired contact points only. The model is built using the con- cept of Virtual Frame(VF), and it is independent from the choice of hand frame and object frame. The performance of the proposed approach is validated on two differ- ent robotic hands, an industrial gripper (4 DOF Barrett hand) and a humanoid hand (16 DOF Allegro hand) to manipulate objects of daily use with complex geometry and various texture (a spray bottle, a tea caddy, a jug and a bunny toy). In the second part of this thesis, we propose an approach to the design of adapta- tion strategy to ensure grasp stability in the presence of physical uncertainties of objects(object weight, friction at contacts and external perturbation). Based on an object-level impedance controller, we first design a grasp stability estimator in the object frame using the grasp experience and tactile sensing. Once a grasp is predicted to be unstable during online execution, the grasp adaptation strategy is triggered to improve the grasp stability, by either changing the stiffness at finger level or relocating the position of one fingertip to a better area.

EPFL2016

Control and Learning of Compliant Manipulation Skills

Klas Jonas Alfred Kronander

EPFL2015

Dynamic Grasp Adaptation

Miao Li

EPFL2016