Calibration of tactile/force sensors for grasping with the PRISMA Hand II


Canbay D. E. , Ferrentino P., Liu H., Moccia R., Pirozzi S., Siciliano B., ...More

IEEE/ASME International Conference on Advanced Intelligent Mechatronics (AIM), ELECTR NETWORK, 12 - 16 July 2021, pp.442-447 identifier identifier

  • Publication Type: Conference Paper / Full Text
  • Doi Number: 10.1109/aim46487.2021.9517508
  • Country: ELECTR NETWORK
  • Page Numbers: pp.442-447
  • Keywords: robotic hand, force/tactile sensors, neural networks models, force regulation, SYNERGY-BASED CONTROL, DESIGN

Abstract

The PRISMA Hand II is a mechanically robust anthropomorphic hand developed at PRISMA Lab, University of Naples Federico II. The hand is highly underactuated, three motors drive 19 joints via elastic tendons. Thanks to its particular mechanical design, the hand can perform not only adaptive grasps but also in-hand manipulation. Each fingertip integrates a tactile/force sensor, based on optoelectronic technology, to provide tactile/force feedback during grasping and manipulation, particularly useful with deformable objects. The paper briefly describes the mechanical design and sensor technology of the hand and proposes a calibration procedure for tactile/force sensors. A comparison between different models of Neural Networks architectures, suitable for sensors calibration, is shown. Experimental tests are provided to choose the optimal tactile sensing suite. Finally, experiments for the regulation of the forces are made to show the effectiveness of calibrated sensors.