Virtual reality support for teleoperation using online grasp planning

Hertkorn, Katharina; Roa, Máximo A.; Brucker, Manuel; Kremer, Philipp; Borst, Christoph

doi:10.1109/iros.2013.6696637

Cited by 9 publications

(11 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As a platform for the proposed method, we use a shared autonomy system consisting of DLR's telepresence robot combined with online grasp planning [2], as described in Section IV-A. However, the only system-dependent assumption made is that the hands are capable of estimating whether each finger is in contact with an object, and the tracking approach should apply equally well to any system that satisfies this requirement.…”

Section: Methodsmentioning

confidence: 99%

“…The shared autonomy setup poses additional challenges to the tracking system, as the human operator can move the arms freely (in contrast to executing a pre-planned motion), potentially bumping the objects or otherwise changing the hand-to-object relative pose, and the new pose is needed by the grasp planner before a new grasp can be planned. Following previous work [2], we use the static scene analysis as an initial guess for the presented tracking method, which, being a local gradient descent method, needs a rough initialization. We perform a quantitative evaluation of the tracking accuracy on real manipulations and show the benefits of the phsyical constraints using a synthetic manipulation as a baseline.…”

Section: Methodsmentioning

confidence: 99%

“…In previous work on the shared autonomy system [2], the object pose was estimated while it lay stationary on the table, not occluded by either hand. When handing the object from one hand to the other, however, the grasp planner needs to know the relative transformation between the grasping hand and an object which is not on the table, is occluded by the fingers of the other hand, and could be moving.…”

Section: Real Experimentsmentioning

confidence: 99%

“…A common solution is to visually track the object within the frame of reference of some camera; the object pose is then related to the hand pose through a transform from the camera frame of reference to the robot base frame of reference followed by a transform from the robot base frame of reference to the hand frame of reference [1], [2]. The latter depends on a calibrated proprioceptive system, while the former depends on a an extrinsic camera calibration.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Depth-based tracking with physical constraints for robot manipulation

Schmidt

Hertkorn

Newcombe

et al. 2015

2015 IEEE International Conference on Robotics and Automation (ICRA)

Self Cite

View full text Add to dashboard Cite

This work integrates visual and physical constraints to perform real-time depth-only tracking of articulated objects, with a focus on tracking a robot's manipulators and manipulation targets in realistic scenarios. As such, we extend DART, an existing visual articulated object tracker, to additionally avoid interpenetration of multiple interacting objects, and to make use of contact information collected via torque sensors or touch sensors. To achieve greater stability, the tracker uses a switching model to detect when an object is stationary relative to the table or relative to the palm and then uses information from multiple frames to converge to an accurate and stable estimate. Deviation from stable states is detected in order to remain robust to failed grasps and dropped objects. The tracker is integrated into a shared autonomy system in which it provides state estimates used by a grasp planner and the controller of two anthropomorphic hands. We demonstrate the advantages and performance of the tracking system in simulation and on a real robot. Qualitative results are also provided for a number of challenging manipulations that are made possible by the speed, accuracy, and stability of the tracking system.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Section: Real Experimentsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Depth-based tracking with physical constraints for robot manipulation

Schmidt

Hertkorn

Newcombe

et al. 2015

2015 IEEE International Conference on Robotics and Automation (ICRA)

Self Cite

View full text Add to dashboard Cite

show abstract

“…ICRs have been successfully used as part of a shared autonomy approach for telemanipulation, where the human decides the hand pose with respect to the object based on the ICR information [4,5]. This work uses the ICRs as part of a fully autonomous motion planner.…”

Section: Chapter 1 Introductionmentioning

confidence: 99%

Integrated grasp and motion planning using independent contact regions

Fontanals

Dang-Vu

Porges

et al. 2014

2014 IEEE-RAS International Conference on Humanoid Robots

Self Cite

View full text Add to dashboard Cite

Traditionally, grasp and arm motion planning are considered as separate tasks. This might lead to problems such as limitations in the grasp possibilities, or unnecessary long times in the solution of the planning problem. This thesis presents an integrated approach that only requires the initial configuration of the robotic arm and the pose of the target object to simultaneously plan a good hand pose and arm trajectory to grasp the object.The planner exploits the concept of independent contact regions to look for the best possible grasp. In this document, two different methods have been considered to search for good end-effector poses. One biases a sampling approach towards favorable regions using principal component analysis, and the other one considers the capabilities of the robotic arm to decide the most promising hand poses. The performance of the methods is evaluated using different scenarios for the humanoid robot SpaceJustin. In order to validate the paths, some scenarios were replicated in the laboratory and the generated paths were executed on the real robot.

show abstract

On Visual Servoing to Improve Performance of Robotic Grasping

Gridseth

Hertkorn

Jägersand

2015

2015 12th Conference on Computer and Robot Vision

Self Cite

View full text Add to dashboard Cite

We introduce image based visual servoing (IBVS) into a shared autonomy grasping system to improve its performance. Visual servoing is a technique that uses visual input to control a dynamic system, such as a robot. Autonomous grasp planning is used to calculate stable grasps to simplify the user control over a robot hand to 1 degree of freedom (DOF) open and close in contrast to controlling every finger. Visual servoing serves to increase the performance of completing the calculated grasp by using visual input to move some of the robot fingers to their grasp points. In this paper, we detail what we have accomplished, the challenges we have faced and what we have learned from them throughout the development of our system.

show abstract

Virtual reality support for teleoperation using online grasp planning

Cited by 9 publications

References 9 publications

Depth-based tracking with physical constraints for robot manipulation

Depth-based tracking with physical constraints for robot manipulation

Integrated grasp and motion planning using independent contact regions

On Visual Servoing to Improve Performance of Robotic Grasping

Contact Info

Product

Resources

About