AUTHOR=Lii Neal Y. , Pereira Aaron , Dietl Julian , Stillfried Georg , Schmidt Annika , Beik-Mohammadi Hadi , Baker Thomas , Maier Annika , Pleintinger Benedikt , Chen Zhaopeng , Elawad Amal , Mentzer  Lauren , Pineault Austin , Reisich Philipp , Albu-Schäffer Alin TITLE=Exodex Adam—A Reconfigurable Dexterous Haptic User Interface for the Whole Hand JOURNAL=Frontiers in Robotics and AI VOLUME=8 YEAR=2022 URL=https://www.frontiersin.org/journals/robotics-and-ai/articles/10.3389/frobt.2021.716598 DOI=10.3389/frobt.2021.716598 ISSN=2296-9144 ABSTRACT=

Applications for dexterous robot teleoperation and immersive virtual reality are growing. Haptic user input devices need to allow the user to intuitively command and seamlessly “feel” the environment they work in, whether virtual or a remote site through an avatar. We introduce the DLR Exodex Adam, a reconfigurable, dexterous, whole-hand haptic input device. The device comprises multiple modular, three degrees of freedom (3-DOF) robotic fingers, whose placement on the device can be adjusted to optimize manipulability for different user hand sizes. Additionally, the device is mounted on a 7-DOF robot arm to increase the user’s workspace. Exodex Adam uses a front-facing interface, with robotic fingers coupled to two of the user’s fingertips, the thumb, and two points on the palm. Including the palm, as opposed to only the fingertips as is common in existing devices, enables accurate tracking of the whole hand without additional sensors such as a data glove or motion capture. By providing “whole-hand” interaction with omnidirectional force-feedback at the attachment points, we enable the user to experience the environment with the complete hand instead of only the fingertips, thus realizing deeper immersion. Interaction using Exodex Adam can range from palpation of objects and surfaces to manipulation using both power and precision grasps, all while receiving haptic feedback. This article details the concept and design of the Exodex Adam, as well as use cases where it is deployed with different command modalities. These include mixed-media interaction in a virtual environment, gesture-based telemanipulation, and robotic hand–arm teleoperation using adaptive model-mediated teleoperation. Finally, we share the insights gained during our development process and use case deployments.