Luke Beaudoin
Research Mentor: Mark Draelos
Mentor Department: Robotics, Engineering
Author(s): Not Available
Session: Session 7 (4:00 PM – 4:50 PM)
Presentation Type: Poster 122
Abstract
Haptic control devices enable intuitive human-robot interaction by allowing users to feel a robot’s contact with its environment, an especially valuable capability in medical applications. One widely used device is the Geomagic Touch, in which users manipulate a floating “penâ€_x009d_ connected through a series of articulated links to input 3D coordinates. Our goal is to develop a digital twin that replicates the device’s pose using joint-angle encoder data, ensuring accurate representation of its state, even in fully digital environments such as virtual reality. We propose a digital twin framework for the Geomagic Touch haptic device, driven by joint angles obtained through the OpenHaptics API. These inputs are applied to a mathematical model to approximate the end-effector’s position and orientation in 3D space. In order to accomplish this, we used existing mechanical design files of the Geomagic Touch to create a nominal representation. Calibration data for the real world coordinates was collected by moving the end effector to 62 preset holes with known space on a precision optical table and recording the modelled position for each position. From this data, three metrics of error were calculated: mean point-to-point distance, mean “ideal positionâ€_x009d_ offset, and an error matrix between expected and modelled positions. This error matrix was used in two ways to correct the model: firstly by using the error matrix to manually offset the end position, and secondly by performing a variable optimization on the model parameters. We believe use of either method would reduce measured points of error. The plan for this project is to create a validated model for the Touch ready for use in medical applications, to a degree appropriate for the resulting degree of accuracy and precision provided by the model.
