Understanding force interactions in tasks for robot programming

  • A collaborative effort between the UW Graphics group, Wisconsin HCI lab and the REACH lab, we explore methods that allow communication of tasks that depend significantly on tactile information such as forces and their utility for robotics applications.
  • Faculty:
  • Michael Zinn – Department of Mechanical Engineering, University of Wisconsin-Madison
  • Michael Gleicher – Department of Computer Science, University of Wisconsin-Madison
  • Bilge Mutlu – Department of Computer Science, University of Wisconsin-Madison
  • Students:   
  • Guru Subramani  – Department of Mechanical Engineering, University of Wisconsin-Madison
  • Danny Rakita  – Department of Computer Science, University of Wisconsin-Madison
  • Funding:
  • UW2020 – WARF Discovery Initiative (UW-Madison)

Recent Work

  • ICRA/RAL 2018 – Recognizing Geometric Constraints in Human Demonstrations using Force and Position Signals Abstract—This paper introduces a method for recognizing geometric constraints from human demonstrations using both position and force measurements. Our key idea is that position information alone is insufficient to determine that a constraint is active and reaction forces must also be considered to correctly distinguish constraints from movements that just happen to follow a particular ...
  • Instrumented Tongs – An input method for tactile manipulations As most interactions with everyday objects involve complicated force interactions, we design and built instrumented tongs that measure dynamic interaction with objects in the environment. This novel approach allows accurate measurement of interactions in a controlled environment while still maintaining dexterity. These tongs measure grasp forces using two optoforce force sensors. We measure the location ...
  • IROS 2017 – Recognizing Actions during Tactile Manipulations through Force Sensing Abstract—In this paper we provide a method for identifying and temporally localizing tactile force actions from measured force signals. Our key idea is to use the continuous wavelet transform (CWT) with the Complex Morlet wavelet to transform force signals into feature vectors amenable to machine learning algorithms. Our method uses these feature vectors to train ...