Object shape reconstruction based on the object manipulation

The International Journal of Advanced Manufacturing Technology, 2017

Tactile sensors are key components for a robot hand system, which are usually used to obtain the object's features. The use of tactile sensors to obtain information from the objects is an open topic of research. In this paper, a new strategy for in-hand extraction of object's properties and to control the interaction forces with robot fingers, mainly based on tactile data, is presented. The scope of this strategy is to grasp and manipulate solid objects, including rigid and soft bodies. Assuming that the hand is in an initial configuration in which the object is grasped, the properties' extraction approach is executed. After the extraction of properties is finished, the object can be classified in regard to a general body listing: rigid body, soft elastic body or soft plastic object. Once the object is classified, for in-hand manipulation tasks, the contact points between the object grasped and the fingers are maintained using the information given by the tactile sensors in order to perform manipulation tasks. Each task is defined by a sequence of basic actions, in which the contact points and applied forces are adapted depending on the action to be performed, and the estimated features for the object. The presented approach tries to imitate the behavior of human beings, in which the applied forces by the fingers are changed when the human estimates the rigidity of a body and when the fingers react to unexpected movements of the object to keep the contact points.

Sensors (Basel, Switzerland), 2018

This work presents a novel and simple approach in the area of manipulation of unknown objects considering both geometric and mechanical constraints of the robotic hand. Starting with an initial blind grasp, our method improves the grasp quality through manipulation considering the three common goals of the manipulation process: improving the hand configuration, the grasp quality and the object positioning, and, at the same time, prevents the object from falling. Tactile feedback is used to obtain local information of the contacts between the fingertips and the object, and no additional exteroceptive feedback sources are considered in the approach. The main novelty of this work lies in the fact that the grasp optimization is performed on-line as a reactive procedure using the tactile and kinematic information obtained during the manipulation. Experimental results are shown to illustrate the efficiency of the approach.

Robotics and Autonomous Systems, 2017

Tactile information has been largely exploited for object recognition with robotic hands but very few approaches have used proprioception alone. In those that do, raw values of joint angles or torques are exploited to train learning algorithms. However, these approaches under-exploit the potential of proprioception, such as its usefulness to estimate the object pose and size. Furthermore, they focus on recognizing individual objects, which increases the amount of data needed to train the algorithms. In this paper, we present an approach based only on joint angles of a robotic hand to generate a shape proprioceptive signature that is invariant to the size and position of the object. Instead of recognizing a specific object from a list, object characteristics useful for its manipulation are extracted. This signature is exploited not only for shape recognition but also for pose estimation. To illustrate the scope of this method, tests are performed on primitive shapes. Results show that the signatures are invariant within large ranges of sizes and poses. Experiments on real hand were carried, and results depicted that the method works similarly in both simulated environment and real applications. A comparison between this two results is made and discussed.

2012 IEEE International Conference on Multisensor Fusion and Integration for Intelligent Systems (MFI), 2012

Proceedings of IEEE International Conference on Robotics and Automation, 1996

In this paper we propose an approach -we call it a "Blind Man's" approach -to shape description in which tactile information is sensed from the fingertips of a dexterous hand. Using this contact information, we investigate two complementary methods for curvature estimation. The first method is based on rolling one finger to estimate curvature at a point on the surface. We use Montana's equations for estimating curvature at a point using simulations and analyze the sensitivity of the approach to noise. The second method uses multiple fingers to slide along a surface while sensing contact points and surface normals. We present a method to extract the shape properties of a patch obtained by fitting a B-spline surface to this multi-fingered sweep across the surface of the object. The method enables us to extract higher level shape information based on the curvature properties of patch.

Robotics and Autonomous Systems, 2017

The field of in-hand robot manipulation of deformable objects is an open and key issue for the next-coming robots. The complexity of physics in a soft object and the huge variety of the deformations of its deformable shape, which can be caused when a robot grasps and manipulates it, makes necessary to develop an adaptable and agile framework for these tasks. Many research works have been proposed to control the manipulation tasks using a model of the manipulated object. Despite these techniques are precise to model the deformations, they are time consuming and, using them in real environments is almost impossible because of the large amount of objects which the robot could find. In this paper, we propose a model-independent framework to control the movements of the fingers of the hands while the robot executes manipulation tasks with deformable objects. This technique is based on tactile images which are obtained as a common interface for different tactile sensors, and uses a servo-tactile control to stabilize the grasping points, avoid sliding and adapt the contacts' configuration regarding to position and magnitude of the applied force. Tactile images are obtained using a combination of dynamic Gaussians, which allows the creation of a common representation for tactile data given by different sensors with different technologies and resolutions. The framework was tested on different manipulation tasks which include complex deformations of the objects, and without using a model of them.

When a real or virtual object is picked up using a robotic arm, the strain, e.g., the pattern of deformation, on the object is unclear. Tactile sensing is key to autonomous robotic interaction with the world. This disclosure describes techniques to determine a contour map of the strain on an object being picked up by a robotic arm or other prehensile device. Per the techniques, cameras sensitive to light polarization are used to detect the polarization angle of the light being reflected off the picked-up object. The polarization angle of light reflecting off a given point on the surface is used to infer various object reactions to the gripping pressure, e.g., the contour map of the deformation of the object, etc.

Proceedings of SPIE, 2014

Haptic perception remains a grand challenge for artificial hands. Dexterous manipulators could be enhanced by "haptic intelligence" that enables identification of objects and their features via touch alone. Haptic perception of local shape would be useful when vision is obstructed or when proprioceptive feedback is inadequate, as observed in this study. In this work, a robot hand outfitted with a deformable, bladder-type, multimodal tactile sensor was used to replay four human-inspired haptic "exploratory procedures" on fingertip-sized geometric features. The geometric features varied by type (bump, pit), curvature (planar, conical, spherical), and footprint dimension (1.25-20 mm). Tactile signals generated by active fingertip motions were used to extract key parameters for use as inputs to supervised learning models. A support vector classifier estimated order of curvature while support vector regression models estimated footprint dimension once curvature had been estimated. A distal-proximal stroke (along the long axis of the finger) enabled estimation of order of curvature with an accuracy of 97%. Best-performing, curvature-specific, support vector regression models yielded R 2 values of at least 0.95. While a radial-ulnar stroke (along the short axis of the finger) was most helpful for estimating feature type and size for planar features, a rolling motion was most helpful for conical and spherical features. The ability to haptically perceive local shape could be used to advance robot autonomy and provide haptic feedback to human teleoperators of devices ranging from bomb defusal robots to neuroprostheses.

Proceedings of International Conference on Robotics and Automation, 1997

In this paper we propose an approach for investigating the shape properties of objects with curved surfaces using tactile sensing by a dexterous robot hand. Our proposed Exploratory Procedure (EP) uses multiple fingers to slide along a surface while sensing contact points and surface normals. Our approach incorporates the surface normal information into a B-spline surface fit of the data gathered from this EP. This additional surface normal information was found to improve the approximation of the surface. The shape properties of this B-spline surface are analyzed to recognize the shape of the object. A confidence estimate of the estimated shape is also calculated.