Global Localization of Objects via Touch

Petrovskaya, Anna; Khatib, Oussama

doi:10.1109/tro.2011.2138450

Cited by 132 publications

(106 citation statements)

References 32 publications

Supporting

Mentioning

106

Contrasting

Order By: Relevance

“…Petrovskaya and Khatib [11] applied a particle filter to the problem of estimating the 6 DOF pose of an object rigidly attached to a robot arm, and proposed the use of scaling series to improve the efficiency of the estimation. More recently Martinez-Hernandez et al [12] demonstrated a contour following technique to reconstruct the 2D shape of an object also rigidly attached to a table.…”

Section: Introductionmentioning

confidence: 99%

Shape and pose recovery from planar pushing

Leonard

Rodríguez

2015

2015 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)

View full text Add to dashboard Cite

Abstract-Tactile exploration refers to the use of physical interaction to infer object properties. In this work, we study the feasibility of recovering the shape and pose of a movable object from observing a series of contacts. In particular, we approach the problem of estimating the shape and trajectory of a planar object lying on a frictional surface, and being pushed by a frictional probe. The probe, when in contact with the object, makes observations of the location of contact and the contact normal.Our approach draws inspiration from the SLAM problem, where noisy observations of the location of landmarks are used to reconstruct and locate a static environment. In tactile exploration, analogously, we can think of the object as a rigid but moving environment, and of the pusher as a sensor that reports contact points on the boundary of the object.A key challenge to tactile exploration is that, unlike visual feedback, sensing by touch is intrusive in nature. The object moves by the action of sensing. In the 2D version of the problem that we study in this paper, the well understood mechanics of planar frictional pushing provides a motion model that plays the role of odometry. The conjecture we investigate in this paper is whether the models of frictional pushing are sufficiently descriptive to simultaneously estimate the shape and pose of an object from the cumulative effect of a sequence of pushes.

show abstract

Section: Introductionmentioning

confidence: 99%

Shape and pose recovery from planar pushing

Leonard

Rodríguez

2015

2015 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)

View full text Add to dashboard Cite

show abstract

“…As in [4,8,10], we adopt the simplifying approximation of supposing both measurements z P and z n independent if conditioned on x, φ c and z CD . Therefore, the joint measurement probability can be expressed as in Equation (11).…”

Section: The Robot Makes Contact With the Objectmentioning

confidence: 99%

“…To the best of our knowledge, literature examples presented case studies with polygonal meshes up to about 100 faces [8,10]. Here, we contribute to the state-of-the-art localizing objects up to industrial complexity.…”

Section: Localization Of Objects Of Increasing Complexitymentioning

confidence: 99%

DOF Decoupling Task Graph Model: Reducing the Complexity of Touch-Based Active Sensing

2015

View full text Add to dashboard Cite

This article presents: (i) a formal, generic model for active sensing tasks; (ii) the insight that active sensing actions can very often be searched on less than six-dimensional configuration spaces (bringing an exponential reduction in the computational costs involved in the search); (iii) an algorithm for selecting actions explicitly trading off information gain, execution time and computational cost; and (iv) experimental results of touch-based localization in an industrial setting. Generalizing from prior work, the formal model represents an active sensing task by six primitives: configuration space, information space, object model, action space, inference scheme and action-selection scheme; prior work applications conform to the model as illustrated by four concrete examples. On top of the mentioned primitives, the task graph is then introduced as the relationship to represent an active sensing task as a sequence of low-complexity actions defined over different configuration spaces of the object. The presented act-reason algorithm is an action selection scheme to maximize the expected information gain of each action, explicitly constraining the time allocated to compute and execute the actions. The experimental contributions include localization of objects with: (1) a force-controlled robot equipped with a spherical touch probe; (2) a geometric complexity of the to-be-localized objects up to industrial relevance; (3) an initial uncertainty of (0.4 m, 0.4 m, 2π); and (4) a configuration Robotics 2015, 4 142 of act-reason to constrain the allocated time to compute and execute the next action as a function of the current uncertainty. Localization is accomplished when the probability mass within a 5-mm tolerance reaches a specified threshold of 80%. Four objects are localized with final {mean; standard-deviation} error spanning from {0.0043 m; 0.0034 m} to {0.0073 m; 0.0048 m}.

show abstract

“…It is viewed as potentially beneficial for manipulation because it provides reliable geometric information in the last centimeter before contact. The disadvantage of using tactile sensing for collecting local geometric information (as in [5], [6]) is that contacting the object may displace it-an outcome that is particularly likely when the object's geometry is initially uncertain. An advantage of pretouch over vision is that pretouch is not subject to the problem of the hand occluding the camera, because the sensor is integrated into the hand; another is that there are no camera-to-hand registration errors since the sensor is in the coordinate frame of the hand.…”

Section: A Robotic Pretouch Sensingmentioning

confidence: 99%

Seashell effect pretouch sensing for robotic grasping

Jiang

Smith

2012

2012 IEEE International Conference on Robotics and Automation

View full text Add to dashboard Cite

Abstract-This paper introduces seashell effect pretouch sensing, and demonstrates application of this new sensing modality to robot grasp control, and also to robot grasp planning."Pretouch" refers to sensing modalities that are intermediate in range between tactile sensing and vision. The novel pretouch technique presented in this paper is effective on materials that prior pretouch techniques fail on. Seashell effect pretouch is inspired by the phenomenon of "hearing the sea" when a seashell is held to the ear, and in particular relies on the observation that the "sound of the sea" changes as the distance from the seashell to the head varies. To turn the familiar seashell effect into a sensor for robotic manipulation, a cavity and microphone was built into a robot finger. The sensor detects changes in the spectrum of ambient noise that occur when the finger approaches an object. Environmental noise is amplified most (attenuated least) at the cavity's resonant frequency, which changes as the cavity approaches an object.After introducing the sensing modality and characterizing its performance, the paper describes experiments performed with prototype sensors integrated into a Willow Garage PR2 robot's gripper. We explore two primary applications: (1) reactive grasp control and (2) pretouch-assisted grasp planning. For reactive grasp control, we demonstrate the advantage of pretouch over pressure sensing for sensing objects that are too compliant for the PR2's tactile sensors to detect during the grasp execution. We demonstrate the benefit of seashell effect pretouch over prior pretouch methods by demonstrating that the new method succeeds with materials on which prior methods fail. For pretouch-assisted grasp planning, the pointcloud from the PR2's texture-projection / narrow stereo camera is augmented with additional points collected by recording the position of the robot's end effector when the pretouch sensor detects the object. This method can compensate for points that are missing from the pointcloud either because of depth camera failure or occlusion. For this application, pretouch is advantageous over contact sensing because contact sensing tends to unintentionally displace the object.

show abstract

Global Localization of Objects via Touch

Cited by 132 publications

References 32 publications

Shape and pose recovery from planar pushing

Shape and pose recovery from planar pushing

DOF Decoupling Task Graph Model: Reducing the Complexity of Touch-Based Active Sensing

Seashell effect pretouch sensing for robotic grasping

Contact Info

Product

Resources

About