This website requires JavaScript.

Learning Haptic-based Object Pose Estimation for In-hand Manipulation with Underactuated Robotic Hands

Osher AzulayInbar Ben-DavidAvishai Sintov
Jul 2022
摘要
Unlike traditional robotic hands, underactuated compliant hands arechallenging to model due to inherent uncertainties. Consequently, poseestimation of a grasped object is usually performed based on visual perception.However, visual perception of the hand and object can be limited in occluded orpartly-occluded environments. In this paper, we aim to explore the use ofhaptics, i.e., kinesthetic and tactile sensing, for pose estimation and in-handmanipulation with underactuated hands. Such haptic approach would mitigateoccluded environments where line-of-sight is not always available. We put anemphasis on identifying the feature state representation of the system thatdoes not include vision and can be obtained with simple and low-cost hardware.For tactile sensing, therefore, we propose a low-cost and flexible sensor thatis mostly 3D printed along with the finger-tip and can provide implicit contactinformation. Taking a two-finger underactuated hand as a test-case, we analyzethe contribution of kinesthetic and tactile features along with variousregression models to the accuracy of the predictions. Furthermore, we propose aModel Predictive Control (MPC) approach which utilizes the pose estimation tomanipulate objects to desired states solely based on haptics. We have conducteda series of experiments that validate the ability to estimate poses of variousobjects with different geometry, stiffness and texture, and show manipulationto goals in the workspace with relatively high accuracy.
展开全部
图表提取

暂无人提供速读十问回答

论文十问由沈向洋博士提出,鼓励大家带着这十个问题去阅读论文,用有用的信息构建认知模型。写出自己的十问回答,还有机会在当前页面展示哦。

Q1论文试图解决什么问题?
Q2这是否是一个新的问题?
Q3这篇文章要验证一个什么科学假设?
0
被引用
笔记
问答