Tactile Internet: with Human-in-the-Loop
()
About this ebook
Tactile Internet with Human-in-the-Loop describes the change from the current Internet, which focuses on the democratization of information independent of location or time, to the Tactile Internet, which democratizes skills to promote equity that is independent of age, gender, sociocultural background or physical limitations. The book promotes the concept of the Tactile Internet for remote closed-loop human-machine interaction and describes the main challenges and key technologies. Current standardization activities in the field for IEEE and IETF are also described, making this book an ideal resource for researchers, graduate students, and industry R&D engineers in communications engineering, electronic engineering, and computer engineering.
- Provides a comprehensive reference that addresses all aspects of the Tactile Internet – technologies, engineering challenges, use cases and standards
- Written by leading researchers in the field
- Presents current standardizations surrounding the IETF and the IEEE
- Contains use cases that illustrate practical applications
Read more from Frank H. P. Fitzek
Qt for Symbian Rating: 0 out of 5 stars0 ratingsMobile Clouds: Exploiting Distributed Resources in Wireless, Mobile and Social Networks Rating: 0 out of 5 stars0 ratings
Related to Tactile Internet
Related ebooks
Computer Vision for Microscopy Image Analysis Rating: 0 out of 5 stars0 ratingsDeep Learning on Edge Computing Devices: Design Challenges of Algorithm and Architecture Rating: 0 out of 5 stars0 ratingsHuman-Machine Shared Contexts Rating: 0 out of 5 stars0 ratingsSmart Home Technologies and Services for Geriatric Rehabilitation Rating: 0 out of 5 stars0 ratingsOptical Fibres: EPO Applied Technology Series Rating: 0 out of 5 stars0 ratingsVLSI Electronics: Microstructure Science Rating: 0 out of 5 stars0 ratingsQuantum Well Lasers Rating: 0 out of 5 stars0 ratingsMultimodal Behavior Analysis in the Wild: Advances and Challenges Rating: 0 out of 5 stars0 ratingsAdvanced Digital Signal Processing and Noise Reduction Rating: 5 out of 5 stars5/5Understanding Microelectronics: A Top-Down Approach Rating: 0 out of 5 stars0 ratingsMobile Computing Deployment and Management: Real World Skills for CompTIA Mobility+ Certification and Beyond Rating: 0 out of 5 stars0 ratingsHandbook of Blind Source Separation: Independent Component Analysis and Applications Rating: 0 out of 5 stars0 ratingsSatellite Signal Propagation, Impairments and Mitigation Rating: 0 out of 5 stars0 ratingsMathematical Theory of Connecting Networks and Telephone Traffic Rating: 0 out of 5 stars0 ratingsHandbook of Human Centric Visualization Rating: 0 out of 5 stars0 ratingsCarrier Ethernet Services A Clear and Concise Reference Rating: 0 out of 5 stars0 ratingsCognitive Big Data Intelligence with a Metaheuristic Approach Rating: 0 out of 5 stars0 ratingsMedical Engineering: Projections for Health Care Delivery Rating: 0 out of 5 stars0 ratingsApplications of Nonlinear Fiber Optics Rating: 0 out of 5 stars0 ratingsVLSI in Medicine Rating: 5 out of 5 stars5/5Ocean Solutions, Earth Solutions Rating: 0 out of 5 stars0 ratingsError-Correction on Non-Standard Communication Channels Rating: 0 out of 5 stars0 ratingsSystems Approach to Appropriate Technology Transfer: Proceedings of the IFAC Symposium, Vienna, Austria, 21-23 March 1983 Rating: 0 out of 5 stars0 ratingsVLSI Implementations for Image Communications Rating: 0 out of 5 stars0 ratingsHandbook of Neural Computing Applications Rating: 5 out of 5 stars5/5Cognitive Radio Technology Rating: 3 out of 5 stars3/5
Networking For You
Mike Meyers' CompTIA Network+ Certification Passport, Sixth Edition (Exam N10-007) Rating: 1 out of 5 stars1/5Network+ Study Guide & Practice Exams Rating: 4 out of 5 stars4/5A Beginner's Guide to Ham Radio Rating: 0 out of 5 stars0 ratingsCybersecurity: The Beginner's Guide: A comprehensive guide to getting started in cybersecurity Rating: 5 out of 5 stars5/5Cisco Networking All-in-One For Dummies Rating: 4 out of 5 stars4/5Networking For Dummies Rating: 5 out of 5 stars5/5AWS Certified Cloud Practitioner Study Guide: CLF-C01 Exam Rating: 5 out of 5 stars5/5CCNA Certification Study Guide, Volume 2: Exam 200-301 Rating: 0 out of 5 stars0 ratingsPractical Ethical Hacking from Scratch Rating: 5 out of 5 stars5/5Home Networking Do-It-Yourself For Dummies Rating: 4 out of 5 stars4/5Networking All-in-One For Dummies Rating: 5 out of 5 stars5/5Linux Bible Rating: 0 out of 5 stars0 ratingsProgramming Arduino: Getting Started with Sketches Rating: 4 out of 5 stars4/5CompTIA Network+ Practice Tests: Exam N10-008 Rating: 0 out of 5 stars0 ratingsCompTIA Network+ Certification Guide (Exam N10-008): Unleash your full potential as a Network Administrator (English Edition) Rating: 0 out of 5 stars0 ratingsThe Compete Ccna 200-301 Study Guide: Network Engineering Edition Rating: 5 out of 5 stars5/5Amazon Web Services (AWS) Interview Questions and Answers Rating: 5 out of 5 stars5/5MCA Microsoft Certified Associate Azure Administrator Study Guide: Exam AZ-104 Rating: 0 out of 5 stars0 ratingsRaspberry Pi Electronics Projects for the Evil Genius Rating: 3 out of 5 stars3/5The Windows Command Line Beginner's Guide: Second Edition Rating: 4 out of 5 stars4/5Microsoft Azure For Dummies Rating: 0 out of 5 stars0 ratingsConcise and Simple Guide to IP Subnets Rating: 5 out of 5 stars5/5SharePoint For Dummies Rating: 0 out of 5 stars0 ratingsCisco Packet Tracer for Beginners Rating: 5 out of 5 stars5/5Applied Network Security Monitoring: Collection, Detection, and Analysis Rating: 3 out of 5 stars3/5Emergency Preparedness and Off-Grid Communication Rating: 0 out of 5 stars0 ratingsEarning Money through Crypto Currency Airdrops, Faucets, Cloud Mining, Online Trading and Online Advertisements Rating: 0 out of 5 stars0 ratingsActive Directory with PowerShell Rating: 4 out of 5 stars4/5
Reviews for Tactile Internet
0 ratings0 reviews
Book preview
Tactile Internet - Frank H. P. Fitzek
Tactile Internet
with Human-in-the-Loop
First edition
Frank H.P. Fitzek
Technische Universität Dresden, Dresden, Germany
Shu-Chen Li
Technische Universität Dresden, Dresden, Germany
Stefanie Speidel
National Center for Tumor Diseases, Partner Site Dresden, Division of Translational Surgical Oncology, Dresden, Germany
Thorsten Strufe
Karlsruhe Institute of Technology, Karlsruhe, Germany
Meryem Simsek
University of California, Berkeley, International Computer Science Institute, Berkeley, CA, United States
Martin Reisslein
Arizona State University, School of Electrical, Computer and Energy Engineering, Tempe, AZ, United States
publogoTable of Contents
Cover image
Title page
Copyright
List of contributors
About the editors
Preface
Acknowledgments
Acronyms
Chapter 1: Tactile Internet with Human-in-the-Loop: New frontiers of transdisciplinary research
Abstract
1.1. Motivation and vision of TaHiL
1.2. Research objectives to meet the challenges of TaHiL
1.3. A synergistic research program
1.4. Research outreaches and societal impacts
1.5. Conclusion and outlook
Bibliography
Part 1: Domains of applications
Introduction
Outline
Chapter 2: Surgical assistance and training
Abstract
2.1. Introduction
2.2. Human-to-machine: Modeling surgical skills
2.3. Machine-to-human: Surgical training
2.4. Human–machine collaboration: Context-aware assistance
2.5. Conclusion and outlook
Bibliography
Chapter 3: Human–robot cohabitation in industry
Abstract
3.1. Introduction
3.2. Model-based cobotic cells
3.3. Tactile robots in the Tactile Internet
3.4. Embedded hardware for robotics
3.5. Synergistic links
3.6. Conclusion and outlook
Bibliography
Chapter 4: Internet of Skills
Abstract
4.1. Aims of the Internet of Skills
4.2. State-of-the-art research: Skill learning and technology
4.3. Key requirements and challenges in designing skill learning with TaHiL technology
4.4. Beyond the state-of-the-art approach: Interdisciplinary collaboration
4.5. Conclusion and outlook
Bibliography
Part 2: Key technology breakthroughs
Introduction
Outline
Chapter 5: Haptic codecs for the Tactile Internet
Abstract
5.1. Scope of haptic codecs
5.2. State-of-the-art research and technology
5.3. Key challenges
5.4. Approaches addressing challenges and beyond the state of the art
5.5. Synergistic links
5.6. Conclusion and outlook
Bibliography
Chapter 6: Intelligent networks
Abstract
6.1. Introduction and motivation
6.2. Evolution of communication networks
6.3. TaHiL communication concept
6.4. Architecture discussion
6.5. TaHiL testbeds
6.6. Synergy and collaboration
6.7. Conclusion and outlook
Bibliography
Chapter 7: Augmented perception and interaction
Abstract
7.1. Milestones for building Human-in-the-Loop systems
7.2. State-of-the-art research and technology
7.3. Identified key challenges of current research and technology
7.4. Research within TaHiL
7.5. Conclusion and outlook
Bibliography
Chapter 8: Human-inspired models for tactile computing
Abstract
8.1. Motivation and aims
8.2. Neuroscientific insights into human decision-making
8.3. Human-inspired learning
8.4. Synergetic links
8.5. Conclusion and outlook
Bibliography
Part 3: Fundamental challenges
Introduction
Outline
Chapter 9: Human perception and neurocognitive development across the lifespan
Abstract
9.1. Introduction: Multisensory perception is the gateway for interactions
9.2. State-of-the-art research on multisensory perception
9.3. Outstanding challenges in current research
9.4. Beyond the state of the art: synergistic research across disciplines
9.5. Conclusion and outlook
Bibliography
Chapter 10: Sensors and actuators
Abstract
10.1. Sensors and actuators of the future
10.2. State of the art
10.3. Key challenges
10.4. Beyond the state-of-art approaches
10.5. Synergistic links
10.6. Conclusion and outlook
Bibliography
Chapter 11: Communications and control
Abstract
11.1. Motivation
11.2. Research in the field of control
11.3. Research in the field of communications
11.4. Conclusion and outlook
Bibliography
Chapter 12: Tactile electronics
Abstract
12.1. Goals
12.2. State of the art
12.3. Research challenges
12.4. Research approaches
12.5. Collaboration
12.6. Conclusion and outlook
Bibliography
Chapter 13: Tactile computing: Essential building blocks for the Tactile Internet
Abstract
13.1. Introduction
13.2. Safe and secure infrastructure
13.3. World capturing and modeling
13.4. Scalable computation
13.5. Context-adaptive software for the Tactile Internet
13.6. Self-explanation for Tactile Internet applications
13.7. Conclusion and outlook
Bibliography
Part 4: Technological standards and the public
Introduction
Outline
Chapter 14: Traces for the Tactile Internet: Architecture, concepts, and evaluations
Abstract
14.1. Introduction
14.2. Tactile traces generic system overview
14.3. Tactile trace content examples
14.4. Application scenarios
14.5. Conclusion and outlook
Bibliography
Chapter 15: Tactile Internet standards of the IEEE P1918.1 Working Group
Abstract
15.1. Introduction
15.2. Definition of the Tactile Internet
15.3. IEEE P1918.1 Tactile Internet Standards Working Group
15.4. IEEE P1918.1 architecture
15.5. IEEE P1918.1 use cases
15.6. IEEE P1918.1 haptic codecs
15.7. Conclusion and outlook
Bibliography
Chapter 16: Public opinion and the Tactile Internet
Abstract
16.1. Introduction
16.2. Theoretical background
16.3. Survey on the public opinion regarding the Tactile Internet
16.4. Survey results
16.5. Discussion
16.6. Conclusion and outlook
Bibliography
Bibliography
Bibliography
Index
Copyright
Academic Press is an imprint of Elsevier
125 London Wall, London EC2Y 5AS, United Kingdom
525 B Street, Suite 1650, San Diego, CA 92101, United States
50 Hampshire Street, 5th Floor, Cambridge, MA 02139, United States
The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, United Kingdom
Copyright © 2021 Elsevier Inc. All rights reserved.
No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Details on how to seek permission, further information about the Publisher's permissions policies and our arrangements with organizations such as the Copyright Clearance Center and the Copyright Licensing Agency, can be found at our website: www.elsevier.com/permissions.
This book and the individual contributions contained in it are protected under copyright by the Publisher (other than as may be noted herein).
Notices
Knowledge and best practice in this field are constantly changing. As new research and experience broaden our understanding, changes in research methods, professional practices, or medical treatment may become necessary.
Practitioners and researchers must always rely on their own experience and knowledge in evaluating and using any information, methods, compounds, or experiments described herein. In using such information or methods they should be mindful of their own safety and the safety of others, including parties for whom they have a professional responsibility.
To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume any liability for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions, or ideas contained in the material herein.
Library of Congress Cataloging-in-Publication Data
A catalog record for this book is available from the Library of Congress
British Library Cataloguing-in-Publication Data
A catalogue record for this book is available from the British Library
ISBN: 978-0-12-821343-8
For information on all Academic Press publications visit our website at https://www.elsevier.com/books-and-journals
Publisher: Mara Conner
Acquisitions Editor: Tim Pitts
Editorial Project Manager: Andrea Gallego Ortiz
Production Project Manager: Nirmala Arumugam
Designer: Matthew Limbert
Typeset by VTeX
List of contributors
Gökhan Akgün Technische Universität Dresden, Dresden, Germany
Ercan Altinsoy Technische Universität Dresden, Dresden, Germany
Uwe Aßmann Technische Universität Dresden, Dresden, Germany
Christel Baier Technische Universität Dresden, Dresden, Germany
Tina Bobbe Technische Universität Dresden, Dresden, Germany
Karlheinz Bock Technische Universität Dresden, Dresden, Germany
Sebastian Bodenstedt National Center for Tumor Diseases, Partner Site Dresden, Dresden, Germany
Juan A. Cabrera G. Technische Universität Dresden, Dresden, Germany
Lingyun Chen Technical University of Munich, Munich, Germany
Chokri Cherif Technische Universität Dresden, Dresden, Germany
Darío Cuevas Rivera Technische Universität Dresden, Dresden, Germany
Raimund Dachselt Technische Universität Dresden, Dresden, Germany
Zaher Dawy American University of Beirut, Beirut, Lebanon
Annika Dix Technische Universität Dresden, Dresden, Germany
Clemens Dubslaff Technische Universität Dresden, Dresden, Germany
Sebastian Ebert Technische Universität Dresden, Dresden, Germany
Mohamad Eid New York University Abu Dhabi, Abu Dhabi, United Arab Emirates
Frank Ellinger Technische Universität Dresden, Dresden, Germany
Sven Engesser Technische Universität Dresden, Dresden, Germany
Gerhard P. Fettweis Technische Universität Dresden, Dresden, Germany
Christof W. Fetzer Technische Universität Dresden, Dresden, Germany
Frank H.P. Fitzek Technische Universität Dresden, Dresden, Germany
Norman Franchi Technische Universität Dresden, Dresden, Germany
Isabel Funke National Center for Tumor Diseases, Partner Site Dresden, Dresden, Germany
Eva Goebel Technische Universität Dresden, Dresden, Germany
Diana Göhringer Technische Universität Dresden, Dresden, Germany
Dominik Grzelak Technische Universität Dresden, Dresden, Germany
Başak Güleçyüz Technical University of Munich, Munich, Germany
Sami Haddadin Technical University of Munich, Munich, Germany
Lutz M. Hagen Technische Universität Dresden, Dresden, Germany
Simon Hanisch Technische Universität Dresden, Dresden, Germany
Ardhi Putra Pratama Hartono Technische Universität Dresden, Dresden, Germany
Rania Hassen Technical University of Munich, Munich, Germany
Adamantini Hatzipanayioti Technische Universität Dresden, Dresden, Germany
Jens R. Helmert Technische Universität Dresden, Dresden, Germany
Diego Hidalgo Technical University of Munich, Munich, Germany
Oliver Holland Advanced Wireless Technology Group, Ltd., London, United Kingdom
Thomas Hulin German Aerospace Center (DLR), Oberpfaffenhofen, Germany
Sebastian A.W. Itting Technische Universität Dresden, Dresden, Germany
Lars Johannsmeier Technical University of Munich, Munich, Germany
Stefan J. Kiebel Technische Universität Dresden, Dresden, Germany
Konstantin Klamka Technische Universität Dresden, Dresden, Germany
Stefan Köpsell Technische Universität Dresden, Dresden, Germany
Jens Krzywinski Technische Universität Dresden, Dresden, Germany
Vincent Latzko Technische Universität Dresden, Dresden, Germany
Simone Lenk
Technische Universität Dresden, Dresden, Germany
Fraunhofer-Gesellschaft, Dresden, Germany
Shu-Chen Li Technische Universität Dresden, Dresden, Germany
Jakub Limanowski Technische Universität Dresden, Dresden, Germany
Tianfang Lin Technische Universität Dresden, Dresden, Germany
Yun Lu Technische Universität Dresden, Dresden, Germany
Lisa-Marie Lüneburg Technische Universität Dresden, Dresden, Germany
Christian Mayr Technische Universität Dresden, Dresden, Germany
Sebastian Merchel Technische Universität Dresden, Dresden, Germany
Johannes Mey Technische Universität Dresden, Dresden, Germany
Annett Mitschick Technische Universität Dresden, Dresden, Germany
Jens Müller Technische Universität Dresden, Dresden, Germany
Evelyn Muschter Technische Universität Dresden, Dresden, Germany
Susanne Narciss Technische Universität Dresden, Dresden, Germany
Krzysztof Nieweglowski Technische Universität Dresden, Dresden, Germany
Andreas Nocke Technische Universität Dresden, Dresden, Germany
Andreas Noll Technical University of Munich, Munich, Germany
Luca Oppici Technische Universität Dresden, Dresden, Germany
Sharief Oteafy DePaul University, Chicago, IL, United States
Sebastian Pannasch Technische Universität Dresden, Dresden, Germany
Michael Panzirsch German Aerospace Center (DLR), Oberpfaffenhofen, Germany
Johannes Partzsch Technische Universität Dresden, Dresden, Germany
Dirk Plettemeier Technische Universität Dresden, Dresden, Germany
Ariel Podlubne Technische Universität Dresden, Dresden, Germany
Martin Reisslein Arizona State University, Tempe, AZ, United States
Dominik Rivoir National Center for Tumor Diseases, Partner Site Dresden, Dresden, Germany
Christian Scheunert Technische Universität Dresden, Dresden, Germany
René Schilling Technische Universität Dresden, Dresden, Germany
Anna Schwendicke Technische Universität Dresden, Dresden, Germany
Patrick Seeling Central Michigan University, Mount Pleasant, MI, United States
Merve Sefunç Technische Universität Dresden, Dresden, Germany
Meryem Simsek International Computer Science Institute, Berkeley, CA, United States
Harsimran Singh German Aerospace Center (DLR), Oberpfaffenhofen, Germany
Stefanie Speidel National Center for Tumor Diseases, Partner Site Dresden, Dresden, Germany
Eckehard Steinbach Technical University of Munich, Munich, Germany
Thorsten Strufe Karlsruhe Institute of Technology, Karlsruhe, Germany
Ronald Tetzlaff Technische Universität Dresden, Dresden, Germany
Andreas Traßl Technische Universität Dresden, Dresden, Germany
Andrés Villamil Technische Universität Dresden, Dresden, Germany
Uwe Vogel Fraunhofer-Gesellschaft, Dresden, Germany
Felix von Bechtolsheim Technische Universität Dresden, Dresden, Germany
Jens Wagner Technische Universität Dresden, Dresden, Germany
Lisa Weidmüller Technische Universität Dresden, Dresden, Germany
Jürgen Weitz Technische Universität Dresden, Dresden, Germany
Hans Winger Technische Universität Dresden, Dresden, Germany
Xiao Xu Technical University of Munich, Munich, Germany
Jiajing Zhang Technische Universität Dresden, Dresden, Germany
Sandra Zimmermann Technische Universität Dresden, Dresden, Germany
About the editors
Frank H.P. Fitzek is a Professor and head of the Deutsche Telekom Chair of Communication Networks at Technische Universität Dresden (TUD) coordinating the 5G Lab Germany since 2014. Since 2019 he is a speaker of the German Research Foundation (DFG, Deutsche Forschungsgemeinschaft) Cluster of Excellence Centre for Tactile Internet with Human-in-the-Loop (CeTI). He received his diploma (Dipl.-Ing.) degree in EE from RWTH Aachen, Germany, in 1997 and his Ph.D. (Dr.-Ing.) in EE from the Technical University Berlin, Germany in 2002 and became Adjunct Professor at the University of Ferrara, Italy in the same year. In 2003, he joined Aalborg University as Professor. He has visited various research institutes, including Massachusetts Institute of Technology (MIT), VTT, and Arizona State University. He cofounded several start-up companies since 1999. He received several awards, such as the NOKIA Champion Award and the Nokia Achievement Award. In 2011, he received the SAPERE AUDE research grant from the Danish government, and in 2012 the Vodafone Innovation prize. In 2015, he was awarded the honorary degree Doctor Honoris Causa from Budapest University of Technology and Economics (BUTE).
Shu-Chen Li is a Professor and head of the Chair of Lifespan Developmental Neuroscience at TUD since 2012. She is a speaker of the DFG Cluster of Excellence CeTI since 2019. She received her Ph.D. degree in cognitive psychology from the University of Oklahoma in the USA in 1994. After working as a postdoc at the McGill University in Canada, she continued her research career at the Max Planck Institute for Human Development in Germany for 16 years until she took up the professorship at TUD. From 2006 to 2008, she was also an adjunct professor of the Brain Research Center in the College of Electrical and Computer Engineering at the National Chiao-Tung University in Taiwan. A key aspect of her research focuses on understanding brain mechanisms of neuronal gain control and their implications on age-related differences in perception and cognition across the human life span. For several years she served as the associated editor of Developmental Psychology, one of the flagship journals of the American Psychological Association. She is currently a member of the editorial board of Neuroscience and Biobehavioral Reviews.
Stefanie Speidel is a Professor for Translational Surgical Oncology at the National Center for Tumor Diseases (NCT), Partner Site Dresden, since 2017 and speaker of the DFG Cluster of Excellence CeTI since 2019. She received her Ph.D. (Dr.-Ing.) from Karlsruhe Institute of Technology (KIT) with distinction in 2009 in the context of the research training group Intelligent Surgery (KIT, University of Heidelberg, DKFZ), and led a junior research group Computer-Assisted Surgery from 2012–2016 at KIT. She has been (co)-authoring more than 100 publications and regularly organizes workshops and challenges, including the Endoscopic Vision Challenge@MICCAI as well as the Surgical Data Science workshop. She has been general chair and program chair for a number of international events, including IPCAI and MICCAI conference.
Thorsten Strufe is a Professor for IT Security at Karlsruhe Institute of Technology (KIT), Adjunct Professor for Privacy and Network Security at TUD, a speaker of the DFG Cluster of Excellence CeTI, and director of the Helmholtz Security Labs KASTEL at KIT. His research interests lie in the areas of large distributed systems and social media, with a focus on privacy and resilience. More recently, he has focused on studying user behavior and security in social media, and on ways to provide privacy-friendly and secure social networking services; he is fascinated by protection through decentralization. One of the challenges that drives him is how to create competitive web services and mobile apps without extensive collection of personal information, thus respecting the privacy of their users. To this end, his group measures and analyzes behavioral data on a large scale, develops algorithms and protocols to improve privacy and security, and formally analyzes anonymization networks for making their actual protection against new attacks formally verifiable.
Meryem Simsek is a Senior Research Scientist at the International Computer Science Institute, UC Berkeley, USA. She received her Ph.D. (Dr.-Ing.) from University of Duisburg-Essen on Learning-Based Techniques for Intercell-Interference Coordination in LTE-Advanced Heterogeneous Networks in 2013. Dr. Simsek has initiated and is currently chairing the IEEE Tactile Internet Technical Committee and serves as the Vice Chair for the IEEE P1918.1 Standardization Working Group, which she co-initiated. On the basis of her roles at IEEE, she disseminates and standardizes the achievements of CeTI. She was a recipient of the IEEE Communications Society Fred W. Ellersick Prize in 2015 and the Rising Star in Computer Networking and Communications by N2Women in 2019.
Martin Reisslein is a Professor in the School of Electrical, Computer, and Energy Engineering at Arizona State University (ASU), Tempe, and an external associated investigator with the DFG Cluster of Excellence CeTI, TUD, Germany. He received the Ph.D. in systems engineering from the University of Pennsylvania, Philadelphia, in 1998. He was a post-doctoral researcher with the Fraunhofer FOKUS institute and the Technical University Berlin from 1998 to 2000, when he joined ASU as Assistant Professor.
Preface
Frank H.P. Fitzek; Shu-Chen Li; Stefanie Speidel; Thorsten Strufe; Meryem Simsek; Martin Reisslein
This book is a result of intensive collaborations among the contributors during the period of applying for a grant from the German Research Foundation (DFG, Deutsche Forschungsgemeinschaft) to establish a Cluster of Excellence and from the ongoing research activities during the first year after the Cluster had been successfully established at Technische Universität Dresden (TUD) in 2019. Together with researchers from other participating institutions, including Technical University of Munich (TUM), the Fraunhofer Institutes and the German Aerospace Center (Deutsches Zentrum für Luft- und Raumfahrt), the National Center for Tumor Diseases (Partner Site Dresden) and several international partners, a core team of researchers from five faculties (Electrical Engineering, Mechanical Engineering, Computer Science, Psychology, and Medicine) at TUD launched the Centre for Tactile Internet with Human-in-the-Loop (CeTI) to pursue new frontiers of research to promote disruptive innovations for digitally transmitted human–machine interactions that may revolutionize many aspects of our lives. This new field of transdisciplinary research will tackle a broad spectrum of theoretical, methodological, and technological challenges. In doing so, the emerging research on Tactile Internet with Human-in-the-Loop (TaHiL) will chart new frontiers for basic and applied research in human and engineering sciences to yield breakthroughs for next-generation multimodal, quasi-real-time human–machine interactions in real, virtual, mixed, and remote environments with broad applications in medicine, industry, and digital transformation technologies for daily-life usages.
Advancing the frontiers of science and technology relies on intensive collaborations among established fields of research that in the end may yield transdisciplinary breakthroughs of much broader impacts than the sum of the outputs from the individual disciplines involved. For a large number of researchers from several disciplines to join forces in embarking on disruptive, transdisciplinary research, as in the case of the research on TaHiL, basic understandings about the key principles of the involved fields, common languages, and shared visions need to be developed across the disciplines. Furthermore, synergistic research needs to be systematically structured and interconnected. This book is conceived as a handbook for the research on TaHiL to serve exactly these purposes. The book follows the structure of a synergistic research program with twelve research building blocks that have been established in the Cluster of Excellence CeTI at TUD. The building blocks are hierarchically interconnected, such that together they form a research pyramid (see the synergistic research program introduced in Chapter 1 for details). The respective research aims, approaches, and activities of the twelve building blocks are each covered by a chapter in this book. With the aim to serve as a handbook, representative work in the relevant areas beyond the research and technologies currently pursued in CeTI are also reviewed in the respective chapters.
Following the introduction (Chapter 1), which provides an overview of the research on TaHiL, the twelve chapters are divided into three parts, proceeding from the top to the base of the pyramid of the synergistic research structure. The first part showcases three selected domains of applications, which are robotic-assisted surgery (Chapter 2), human–robot cohabitation in industrial settings (Chapter 3), and Internet of Skills for other daily applications (Chapter 4). These use-cases presented in Part 1 require the key technologies and methods—in particular haptic codecs (Chapter 5), intelligent networks (Chapter 6), augmented perception and interaction (Chapter 7), as well as human-inspired models and computing (Chapter 8)—that are presented in Part 2. The challenges combined have to be tackled by systematically organized integrative research from several disciplines. These target primary research fields are presented in Part 3, which cover basic research on human multisensory perception (Chapter 9), sensors and actuators (Chapter 10), communications and control (Chapter 11), electronics for textile integration (Chapter 12), and tactile computing (Chapter 13). The last part of the book extends to cover cross-cutting topics, such as a digital trace library (Chapter 14) and standardization (Chapter 15) as well as technology transfer and communication to the public (Chapter 16).
This volume can serve as a handbook for the research on TaHiL for students and researchers from several contributing disciplines. For readers who would like to find out what TaHiL is and what applications the research in this new frontier may have, we recommend surveying Chapter 1 and the chapters in Part 1. For researchers who are already acquainted with topics about some aspects of TaHiL, we recommend sampling chapters from Part 1 for the specific applications of interests, and reading through chapters in Part 2, which highlight key areas of technological breakthroughs that require interdisciplinary research. Furthermore, to foster interdisciplinary understanding, the chapters in Part 3 are recommended for students and researchers to gain knowledge about fundamental questions and methods that are important for the research on TaHiL from the perspectives of other disciplines. Last but not least, the chapters in Part 4 address topics on technological standards and public communication, which are also crucial for the success of developing new technologies to serve better human–machine interactions.
This book marks a beginning. We hope it will kindle more interest and attract intensive research attention for the emerging transdisciplinary field of TaHiL. Interested readers are also referred to the CeTI webpage (ceti.one) for research updates.
Dresden, Germany
2021
Acknowledgments
Frank H.P. Fitzek; Shu-Chen Li; Stefanie Speidel; Thorsten Strufe; Meryem Simsek; Martin Reisslein
First of all, we would like to thank the German Research Foundation (DFG, Deutsche Forschungsgemeinschaft).¹ Many results reported in this book are funded by the DFG as part of Germany's Excellence Strategy² in support of the Cluster of Excellence Centre for Tactile Internet with Human-in-the-Loop (CeTI) established at Technische Universität Dresden (TUD).
The editors would like to thank all the authors who have contributed to the different chapters collected in this book. Most of the authors are members of CeTI, including many current CeTI Ph.D. students and postdocs, who have invested significant amounts of time and effort in addition to their regular duties to make this book possible.
Many thanks also go to our international research partners and consultants, such as (i) Muriel Médard from Massachusetts Institute of Technology (MIT), Adam Gazzaley from the University of California San Francisco (UCSF), Uta Noppeney from the Radboud University, and Gene Tsudik from University of California, Irvine (UCI), who supported us while we applied for the excellence initiative funding to establish CeTI or support us in CeTI's Advisory Board.
In alphabetical order, we express deep gratitude to our industrial partners, such as Atlantic Labs, CampusGenius, Deutsche Telekom, Mimetik, and Wandelbots.
We thank our design team, Jens Krzywinski, Tina Bobbe, Lisa Lüneburg, and their associates, for the support in creating designs and graphics for several demonstrators as well as the illustrations presented in the book. Their work not only gives this book a nice touch, but has also helped us to convey CeTI's main ideas of future communication systems to the public over the last years.
We are deeply thankful to Christian Scheunert and Hrjehor Mark for their support in managing the sources and their patience over the last months. It is their achievement to have all the sources of this book pulled together.
The work presented in this book would not have been possible without the endless support of our universities, i.e., Technische Universität Dresden, Technical University of Munich, as well as the Fraunhofer Institutes, the German Aerospace Center (DLR), and the National Center for Tumor Diseases, Partner Site Dresden (NCT).
Dresden, Germany
2021
¹
"Funded by the German Research Foundation (DFG, Deutsche Forschungsgemeinschaft) as part of Germany's Excellence Strategy – EXC 2050/1 – Project ID 390696704 – Cluster of Excellence Centre for Tactile Internet with Human-in-the-Loop (CeTI) of Technische Universität Dresden."
²
A funding program of the Federal Government and the states to strengthen cutting-edge research at universities.
Acronyms
3D Three-dimensional
3GPP 3rd-Generation Partnership Project
5G Fifth Generation
ADC Analog-to-Digital Converter
AFE Analogue Frontend
AG Attribute Grammar
AI Artificial Intelligence
AM Additive Manufacturing
AOP Aspect-Oriented Programming
API Application Programmer Interface
AR Augmented Reality
ARQ Automatic Repeat Requests
ASF Acceleration Sensitivity Function
ASIC Application-Specific Integrated Circuit
ASP Application Service Provider
ASQ Action Sequence
AVB Audio-Video Bridging
BAN Body Area Network
BCH Body Computing Hub
BDD Binary Decision Diagram
BFGS Broyden-Fletcher-Goldfarb-Shanno
CACC Cooperative Adaptive Cruise Control
CATI Computer-Assisted Telephone Interviews
CBR Constant Bit Rate
CBSE Component-based Software Engineering
CELP Code-Excited Linear Prediction
CeTI Centre for Tactile Internet with Human-in-the-Loop
CEW Communication and Early Warning
CI Communication Interruption
CMD Command
CMOS Complementary Metal Oxide Semiconductor
CNN Convolutional Neural Network
CNS Central Nervous System
COM/SDB ComSoc Standards Development Board
COP Context-Oriented Programming
CORA Core Ontologies for Robotics and Automation
CPE Control Plane Entity
CPS Cyber-Physical System
CPU Central Processing Unit
CR Compression Ratio
CT Computed Tomography
DARPP-32 Dopamine- and cAMP-Regulated Neuronal Phosphoprotein
DB Deadband
DC Direct current
DCT Discrete Cosine Transform
DDoS Distributed Denial of Service
DDS Data Distribution Service
DNF Disjunctive Normal Form
DNN Deep Neural Network
DoF Degrees of Freedom
DPMC Dorsal Premotor Cortex
DPR Dynamic Partial Reconfiguration
DSL Domain-Specific Language
DT Detection Threshold
DTAG Deutsche Telekom
DTLS Datagram Transport Layer Security
DVFS Dynamic Voltage and Frequency Scaling
DWT Discrete Wavelet Transform
E2E End-to-End
ECU Electronic Control Units
EEG Electroencephalography
eMBB Enhanced Mobile Broadband
eSAP External Service Access Point
ESE Energy Storage Element
ETSI European Telecommunications Standards Institute
FB Feedback
FDX Fully Depleted Silicon-on-Insulator
fMRI Functional Magnetic Resonance Imaging
FPGA Field Programmable Gate Array
F-RAN Fog Computing based Radio Access Network
FSK Frequency Shift Keying
FSM Finite State Machine
FUNc Network Functional Compression
GN Gateway Node
GNC Gateway Node Controller
GPU Graphics Processing Unit
GRAND Guessing Random Additive Noise Decoding
HCI Human–Computer Interface
HCTG Haptic Codecs Task Group
HDL Hardware Description Language
HIC Haptic Interpersonal Communication
HLS High Level Synthesis
HO Human Operator
HPC High Performance Computing
HPD High Performance Demonstrator
HRTF Head-Related Transfer Function
HSI Human–System Interface
IAT Inter-Arrival Time
IC Integrated Circuit
ICN Information Centric Networks
IEEE Institute of Electrical and Electronics Engineers
IEEE-SA IEEE Standards Association
IMU Inertial Measurement Unit
IoS Internet of Skills
IoT Internet of Things
IP Internet Protocol
IPL Inferior Parietal Lobe
IPS Intraparietal Sulcus
ISDN Integrated Services Digital Network
ISI Inter-Stimulus Interval
ISO International Standards Organization
ISS Input-to-State Stability
ITU-T International Telecommunication Union Standardization Sector
IVR Immersive Virtual Reality
JIGSAWS JHU-ISI Gesture and Skill Assessment Working Set
JND Just Noticeable Difference
JSON Javascript Object Notation
JVM Java Virtual Machine
K Key Technologies and Methods
KPI Key Performance Indicator
LAN Local Area Network
LED Light-Emitting Diode
LGN Lateral Geniculate Nucleus
LNA Low Noise Amplifier
LTE Long Term Evolution
MAC Medium Access Control
MAD Maximally Allowable Delay
MATI Maximum Allowable Transmission Interval
MBCC Model-Based Cobotic Cell
MDE Model-Driven Engineering
MDP Markov Desicion Process
MEC Mobile Edge Cloud
MGC Medial Geniculate Complex
MGD Mini-Batch Stochastic Gradient Descent
MIMO Multiple-Input Multiple-Output
ML Machine Learning
MMT Model Mediated Teleoperation
mMTC Massive Machine Type Communication
MRI Magnetic Resonance Imaging
NAcc Nucleus Accumbens
NC Network Controller
NCS Networked Control System
NesCom New Standards Committee
NFV Network Function Virtualization
NR New Radio
NS Network Slicing
NUI Natural User Interface
OBG Observer-Based Gradient method
OFDM Orthogonal Frequency Division Multiplexing
OLED Organic Light-Emitting Diode
OOK On-Off Keying
OR Operating Room
OS Operating System
OSATS Objective Structured Assessment of Technical Skills
OSM Orthographic Software Modeling
PA Power Amplifier
PAR Project Authorization Request
PC Passivity Controller
PCTL Probabilistic Computation Tree Logic
PDMS Polydimethylsiloxane
PDU Protocol Data Unit
PE Processing Element
PHY Physical Layer
PL Programmable Logic
PLL Phase-Locked Loop
pRRH pico-Remote-Radio-Head
PSNR Peak Signal to Noise Ratio
PTP Precision Time Protocol
PU Polyurethane
QAM Quadrature Amplitude Modulation
QoC Quality-of-Control
QoE Quality-of-Experience
QoP Quality-of-Performance
QoS Quality-of-Service
QPSK Quadrature Phase-Shift Keying
RAG Reference Attribute Grammar
RAM Random-Access Memory
RAN Radio Access Network
RC Reflection Coefficient
RCS Reconfigurable Computing System
RGB Red Green Blue
RGBD Red Green Blue and Depth
RL Reinforcement Learning
RLNC Random Linear Network Coding
RO Robot Operator
ROP Role-Oriented Programming
ROS Robot Operating System
RQs Research Questions
RRI Responsible Research and Innovation
RRM Radio Resource Management
RRSI Rapid Reaction Standardization Initiative
RT Reaction Time
RTOS Real-Time Operating System
SAP Service Access Point
SAS Self-Adaptive System
SAW Spatial Audio Workstation
SDC Silent Data Corruption
SDN Software Defined Network
SE Support Engine
SFA Successive Force Augmentation
SFC Service Function Chaining
SGX Software Guard eXtension
SLP Sparse Linear Prediction
SMPTE Society of Motion Picture and Television Engineers
SMR Signal-to-Mask Ratio
SNc Substantial Nigra Parc Compacta
SNR Signal-to-Noise Ratio
SoA Service-oriented Architecture
SOC Systems-on-Chip
SPIHT Set Partitioning In Hierarchical Trees
SPL Software Product Line
SR Stimulus Response
SR-ARQ Selective Repeat ARQ
STS Superior Temporal Sulcus
SW-ARQ Stop and Wait ARQ
TA Technology Assessment
TADF Thermally Activated Delayed Fluorescence
TaHiL Tactile Internet with Human-in-the-Loop
TAM Technology Acceptance Model
TD Tactile Device
TDPA Time Domain Passivity Approach
TDPA-ER Time Domain Passivity Approach Energy Reflection
TE Tactile Edge
TEE Trusted Execution Environment
TFT Thin Film Transistor
TI Tactile Internet
TIM Tactile Internet Metadata
TLS Transport Layer Security
TNM Tactile Network Manager
ToF Time of Flight
TP Talent Pool
TPU Thermoplastic Polyurethane
TSM Tactile Service Manager
TSN Time Sensitive Network
TSX Transactional Synchronization eXtension
TT Tactile Traces
TUD Technische Universität Dresden
TUM Technical University of Munich
U Use Cases
UE User Equipment
UML Unified Modeling Language
UPE User Plane Entity
URLLC Ultra-Reliable Low-Latency Communication
UTAUT Unified Theory of Acceptance and Use of Technology
V2V/V2I Vehicle-to-Vehicle/Vehicle-to-Infrastructure
V2X Vehicle-to-Any
VCO Voltage Controlled Oscillator
VPL Ventral Posterolateral Nucleus
VPM Ventral Posteromedia Nucleus
VPMC Ventral Premotor Cortex
VR Virtual Reality
VTA Ventral Tegmental Area
WAN Wide Area Network
WFS Wave Field Synthesis
WG Working Group
WiFi Wireless Fidelity
XML eXtensible Markup Language
ZOH Zero-Order Hold
Chapter 1: Tactile Internet with Human-in-the-Loop: New frontiers of transdisciplinary research
Frank H.P. Fitzeka; Shu-Chen Lia; Stefanie Speidelb; Thorsten Strufec aTechnische Universität Dresden, Dresden, Germany
bNational Center for Tumor Diseases, Partner Site Dresden, Dresden, Germany
cKarlsruhe Institute of Technology, Karlsruhe, Germany
What seemingly was often overlooked…is that the human brain itself…is something that is co-shaped by experience…, something that does not operate in an environmental vacuum, but at any moment is subject to environmental constraints and affordances.
– Paul B. Baltes☆
Engineering is a living branch of human activity and its frontiers are by no means exhausted.
– Igor Sikorsky
Above all things expand the frontiers of science: without this the rest counts for nothing.
– Georg C. Lichtenberg
Abstract
The emerging new field of research on Tactile Internet with Human-in-the-Loop (TaHiL) aims to achieve significant breakthroughs to enhance collaborations between humans and machines or—more generally, Cyber-Physical System (CPS)—in real, virtual, and remote environments. The vision of TaHiL is to enable humans to interact with cooperating CPS over intelligent wide-area communication networks to promote equitable access to remote work, medical, learning, social, and recreational opportunities for people of different ages, genders, cultural backgrounds, or physical limitations. Thus reaching far beyond the current state of the art in digitalization and human–machine interaction, the long-term goal of the research on TaHiL is to democratize the access to skills and expertise the same way as the current Internet has democratized the access to information. Capitalizing on recent advancements in the fields of telecommunication, electrical and material engineering, computer science, robotics, psychology, cognitive neuroscience, and medicine, researchers in this new transdisciplinary field are pursuing basic and applied research to (i) advance the understanding about complex dynamics of human goal-directed multisensory perception and action from the psychological, neurocognitive, medical, and computational perspectives; develop novel sensor and actuator technologies that augment the human mind and body; develop fast, bendable, adaptive, and reconfigurable electronics; create intelligent communication networks that connect humans and CPS by continuously adapting and learning to provide low latency, as well as high levels of resilience and security; (v) design new haptic coding schemes to cope with the deluge of information from massive numbers of body sensors; design online learning mechanisms as well as interface solutions for machines and humans to predict and augment each other's actions; and to evaluate the above technological solutions as well as to engage the general public about the potential possibilities and concerns that the new technologies will bring. The research on TaHiL will be essential for diverse applications involving human–machine interactions, including, most notably, in medicine, industry, and the Internet of Skills (IoS). This overview chapter highlights the challenges, directions, programmatic structures and application domains of the new transdisciplinary research endeavor of TaHiL. Key building blocks of TaHiL are presented in details in the 15 subsequent chapters collected in this volume.
Keywords
Age-sensitive design; haptic communication; haptic sensors; machine learning; multimodal feedback; multisensory perception and action; predictive models; robotics; Tactile Internet; user-centered design
1.1 Motivation and vision of TaHiL
In the summer of 1969 the Internet was created by coupling a small number of computer nodes to share files across different locations. Back then a small number of services was available to a small number of experts. Fifty years later the Internet is among the most important global infrastructures worldwide. It has become a key infrastructure that is used by everyone and touching almost every aspect of human daily lives. The current Internet provides the service of democratizing access to information for everybody, independent of location or time (as illustrated in Fig. 1.1). One important enabler for this function of global information access was the introduction of the World Wide Web, which allowed laypersons and computer scientists alike to create contents for distribution and to consume information through and from the Internet, respectively. Focusing on characteristics of the most popular services, such as video streaming, social networking, or web browsing, the current Internet has been optimized for high data rates to facilitate quick access and live consumption during the download. The next-generation Internet, the Tactile Internet (TI), takes these ideas one big step further. It envisions new opportunities and is faced by entirely novel challenges. The Institute of Electrical and Electronics Engineers (IEEE) P1918.1 Tactile Internet Standardization Working Group (http://ti.committees.comsoc.org/)
defines the TI communication platform as: A network or network of networks for remotely accessing, perceiving, manipulating or controlling real, or virtual objects, or processes in perceived real time by humans or machines [1,2]. To transcend the possibilities of just gaining access to information, as we use the Internet today (compare Fig. 1.1), the Human-in-the-Loop approach [3] needs to be thoroughly realized en route to further technological advancements in the TI. The resulting new field of Tactile Internet with Human-in-the-Loop (TaHiL) research aims at democratizing access to skills and expertise to promote equity for people of different age, genders, cultural backgrounds, or physical limitations (see Fig. 1.2). To reach such breakthroughs for bringing digitally transmitted human–machine interactions to a new era, it is indispensable that transdisciplinary research involving researchers from several fields—ranging from psychology, cognitive neuroscience, and medicine to the fields of computer science, electrical, mechanical, and material engineering—needs to be conducted. One such transdisciplinary research center, the Centre for Tactile Internet with Human-in-the-Loop (CeTI), has been recently established at Technische Universität Dresden (TUD).
Fig. 1.1 Nowadays Internet: Democratizing access to information for everybody regardless of location or time.
Fig. 1.2 TaHiL: Democratizing access to skills and expertise to promote equity for people of different ages, genders, cultural backgrounds, or physical limitations.
The Internet today, which is mainly optimized for increased throughput, cannot support TI applications. In this overview chapter, we will highlight several new challenges that need to be overcome to achieve the goals of TaHiL. The subsequent chapters in the book then present exemplified domains of applications that can be enabled through tackling challenges in a structured array of basic research and technological innovations. Here we highlight three frontiers in engineering and human research that still need to be explored and established: (i) a communication network that is optimized for skill (beyond information) transfer and hence supports extremely low latencies and different Quality-of-Service (QoS) support for modalities, such as video, audio, and haptics, novel human–machine interfaces that utilize a large array of sensors and actuators, and systematic understanding of goal-directed human multisensory perception and action, and the impacts of lifespan development and learning on these processes. Albeit establishing the TI as the next-generation infrastructure for global skill transfer is one important factor, the communication network alone falls short of several other challenges. Advanced wearable and adaptive sensors as well as actuators need to be developed as new types of interfaces for communications between humans and machines via the TI. Furthermore, the research on TaHiL has to consider the principles and mechanisms of human goal-oriented multisensory perception and action in people with different ages, learning experiences, and skill levels. Only then will the TI allow broad populations of human users to immerse themselves into virtual, remote, or inaccessible real environments, to exchange skills and expertise; thus create new opportunities and novel ways for people to learn, to work, and to interact (as illustrated in Fig. 1.3).
Fig. 1.3 (left) How may we learn in the future? (middle) How may our work change due to robots? (right) How may TaHiL technologies help the old and the oldest-old in the future?
1.1.1 Skill transfer from humans to machines
Different paths can be chosen to exchange skills among humans and machines. Here we consider a leading example of skill transfer from a human expert to a machine, in this case a standard industry robot (see Fig. 1.4). One way for this type of skill transfer would be to equip a human expert with any kind of human–machine interface, such as a simple remote control (e.g., a game console controller). This solution, albeit simple, has a couple of caveats. Not all human experts are able to operate such a controller, nor do they understand the relationship between the movements and accuracy of the robot.
Fig. 1.4 Skill transfer from humans to machines.
Even replacing the simplistic controller with a more advanced human–machine interface, such as wearables that track and map human behavior directly to the robot, does not solve another issue regarding the scalability of direct remote control. A single machine still needs to be controlled by one human. In scenarios with millions of consumers demand a specific robot skill, the control has to be scaled up and the skill itself has to be transferred to millions of robots to meet the demand of the consumers. Industry has long recognized this problem, and there are standard processes for conveying skills and expertise to industrial robots (as illustrated in Fig. 1.5).
Fig. 1.5 Common industrial view on skill transfer.
A domain expert in such industry use cases explains the necessary expertise and actions a robot has to perform to a computer scientist. This programming expert will then, to the best of his/her understanding, convey these descriptions into the software that is subsequently executed on the industrial robot. Once successful, the software can be deployed to several robots. While this approach is more scalable than remote control, it still has several problems. First, the communication between the human expert and the computer scientist is error prone and often only a best-effort service. Second, the cost factor of the computer scientist is two to six times higher than the cost of the industrial robot. In light of decreasing prices for robots, this ratio will become even higher. Third, completing the skill transfer, potentially in a sequence of several trial-and-error cycles, is rather time-consuming.
To overcome these limitations, a novel approach developed in the field of TaHiL research at TUD has been proposed and implemented (see Fig. 1.6). This innovation builds specifically on the idea of using wearable clothing that is instrumented by sensors, and worn by domain experts. By performing the action routine several times, the activity of the expert is used to train the robot by demonstration through natural human movements. The training sequences recorded by the sensors are evaluated by machine-learning algorithms, which then output the software for the robot automatically. In other words, by combining sensor recordings of human behavior with machine learning, a direct form of human-to-machine demonstration teaching can be established. Such an approach is one of the promising avenues for further research on TaHiL. Indeed, this solution has been spun off in 2018, leading to a start-up company, Wandelbots, that has now developed several products for industry (for example, see Fig. 1.7).
Fig. 1.6 One of the TaHiL approaches for skill transfer using machine learning and demonstration-based teaching developed at TUD and now implemented in the start-up company, Wandelbots.
Fig. 1.7 The CEO of the Volkswagen group testing the demonstration-based robot teaching developed by the TUD spin-off Wandelbots.
This natural demonstration-based teaching is promising. However, conveying the sensor data from the wearables over a communication network to remote robots in real-time is still a big challenge in the research on TaHiL. Future developments along this novel approach may meet the challenges of specific tasks that would require global exchanges of skills, for instance an expert in Europe trains a robot in Tokyo, Japan (see Fig. 1.8). In this scenario, the task of providing the expert with timely multimodal feedback, as it is required for efficient remote training and for giving the expert the feeling of virtually being right next to the remote robot, is particularly challenging. In fact, this is not possible with the Internet technology as it stands today, and derives several challenges ahead of us to realize the TI.
Fig. 1.8 Future scenarios of global skill exchange in the TI.
An overview of the manifold challenges in different research building blocks of TaHiL is shown in Fig. 1.9. The multimodal feedback needs to comprise haptic, in addition to video and audio, information. Each of these information modalities differs in its requirements of bandwidth and latency. It is already clear that video requires more bandwidth at relaxed delay constraints as compared to audio interaction. Furthermore, the properties and requirements for haptic information, so far, have not yet been extensively investigated and are far from been understood. Thus further basic research on haptic information processing in humans and applied research on haptic technologies would be necessary. With respect to latency and reliability of human sensory and perceptual processes, the latencies range from several milliseconds for video, over around 3 ms for audio, to only about 1 ms for haptic information [4] (see also Chapters 5 and 9). Such strict latency requirements create novel challenges, particularly given the laws of physics. Since the speed of light becomes a limiting factor for the possible distance between the human expert and robot for timely real-time remote interaction, as illustrated in the scenarios above. Even without considering the time needed for sensing, encoding, and processing, light travels at around 300 kilometers per millisecond, which limits the distance between the expert and robot to a range that falls substantially short of the requirements of global skill transfer and space communications.
Fig. 1.9 Challenges in global skill exchange via the TI.
Furthermore, human–machine real-time interactions in the form of coworking or training over longer distances will require local predictions of the remote behaviors, which would then also need to be corrected upon reception of the actual remote updates. While modeling a robot in a well-described physical environment without potential error sources is easy and follows the technical specifications of the machine, modeling and predicting human behaviors are significantly harder tasks that depend on many more parameters and are characterized by a huge number of degrees of freedom (see Chapters 9 and 11). These challenges require the Human-in-the-Loop approach [3] to be thoroughly realized on the foundation of human goal-directed perception and action in all aspects and steps of new technological developments.
1.1.2 Skill transfer from machines to humans
So far we have considered one direction of skill transfer, i.e., having the human to teach a machine. But the reverse direction covers scenarios that could be applicable in other use cases. Assuming that the wearables are not only equipped with sensors but also with actuators, learning signals can either be generated live or in advance, and then conveyed to the human user. Taking physical rehabilitation of the elderly as an example, the movements of a remote physiotherapist could be generated online and transmitted to either wearables equipped with actuators that the elderly person wears, or a Cyber-Physical System (CPS) to help performing physiotherapy exercises at home (see Fig. 1.10). The potential application domains for such skill transfer from machines to humans are not limited to health and nursing care, they also cover teaching new skills in schools, at work, or of personal interests, i.e., the broad domain of Internet of Skills (IoS) [5]. Fig. 1.11 depicts two specific examples of this last class of applications that involve training rowing and climbing with specific wearables for detecting and correcting inefficient or potentially harmful movements. These technologies are currently under development in CeTI at TUD.
Fig. 1.10 A scenario of reverse skill transfer from machines to humans: The case of remote health care.
Fig. 1.11 Scenarios of teaching and training humans (see Chapter 4 ) in the specific cases of (left) rowing and (right) climbing.
1.1.3 Skill transfer in holistic settings
There are several other application domains, for which digital skill transfer may also be valuable. Research on TaHiL aims to enable humans and machines to work collaboratively together in multiple learning activities in the future. Besides the aforementioned scenarios, skill transfer among robots of different manufacturers in completely different environments is a further potential application field. Furthermore, digitally mediated learning from human-to-human over long distances or between restrained environments will open disruptive new opportunities for the democratization of expertise and learning opportunities for acquiring various skills. Holistically, global digitally mediated skill exchanges can involve multiple combinations of human-to-machine, machine-to-human, and human-to-human interactions (see Fig. 1.12).
Fig. 1.12 Scenarios of the holistic human–machine skill transfer via the TI.
1.2 Research objectives to meet the challenges of TaHiL
There are several fundamental objectives for the research on TaHiL, which all revolve around the main building blocks of next generation multimodal closed-loop human–machine interactions that take place in the TI in perceived real-time (see Fig. 1.13): the human, who is augmented by large numbers of sensors and actuators that are connected through an intelligent network, cooperates with CPS (e.g., robots and other virtual- or mixed-reality entities) that are equipped with inherent sensors and actuators as well as adaptive learning mechanisms. Such quasi-real-time closed-loop interactions lead to a plethora of multisensory feedback information that has to be conveyed from the human to the machine and back over the same intelligent network, but with different communication characteristics in terms of latency and resilience. Note that the closed-loop human–machine interaction is not limited to one human, robot, or other CPS; instead, the TaHiL concept generalizes to other combinations and extensions that include an arbitrary number of these components in holistic settings of applications.
Fig. 1.13 Conceptual representation of the TaHiL.
This section describes six key research objectives of TaHiL in a logical order. As shown in Fig. 1.13, we start with the Human-in-the-Closed-Loop system reflecting the first objective on human perception and action. In particular, the first objective mainly concerns the modeling and prediction of human goal-directed multisensory perception and action. The second objective focused on human–machine coaugmentation and addresses the novel bendable electronics, sensors, and actuators required for TaHiL. The aforementioned intelligent network is reflected by the third objective, which focuses on developing human–machine networks that can assure real-time communication, storage, and computing for all involved communication elements. The fourth objective concerns learning strategies for humans and machines to learn from and adapt to each other and is therefore called human–machine learning. The fifth objective of human–machine computation targets the computing infrastructure that is necessary for human–machine interactions. The sixth objective, human–machine communication, aims to develop new information theoretical approaches for communication, compression, coding, and control. (See Fig. 1.14.)
Fig. 1.14 Logical derivation of key objectives of the research on TaHiL.
1.2.1 Objective 1: Human perception and action
Model and predict human goal-directed behavior, which entails flexible and dynamic interactions between sensation, multisensory perception, cognition, and action in contexts.
The novel technologies to be developed for human–machine interactions via the TI will create new digital environments for humans to interact with a wide range of CPS, with substantially, if not completely, changed hardware and software interfaces that require extensive multisensory information processing (see Chapter 9, Fig. 9.1). Thus innovative approaches for system and interface designs would need to be developed to optimize the new digitally transmitted closed-loop interaction between humans and machines. To establish the necessary requirements for engineering designs, computational models of flexible, human goal-directed multisensory perception and action will need to be developed (see Chapter 13). These models need to take into account relevant characteristics of individual differences, such as age and levels of expertise. In particular, the processes of human development [6] and aging [7] as well as mechanisms of skill acquisition and mastery can significantly affect the efficiency of various processes underlying goal-directed perception and action at the behavioral and brain levels. Thus models characterized by appropriate human factors are crucial for the development of new algorithms and technologies for human–machine coadaptation, in which goal awareness and action prediction are the prerequisites for smooth interactions. This requires research to go far beyond the current state of understanding. We need to characterize and understand expertise- and age-related differences in key parameters of multisensory integration and delay requirements. Methodologically, psychophysical, and neurocognitive experiments will need to be conducted with large samples of individuals covering wide age ranges and expertise levels. The experiments will need to encompass different sensory modalities (e.g., auditory, visual, and haptic) across an array of perceptual decision and sensorimotor tasks that entail flexible switching between goal sequences or task contexts. Psychophysical and Bayesian active inference models will need to be developed to model and predict expertise- and age-related differences in the complexity and constraints of human goal-directed perception and action (see Chapter 9 for details about research on these topics and experimental studies currently been pursued in CeTI).
1.2.2 Objective 2: Human–machine coaugmentation
Produce wearable peripherals for fast sensing and actuating with multimodal feedback for human perception, cognition, and action based on ultra-small, bendable, stretchable, and ultra-low-power electronic circuits that precisely localize humans and objects in real-time.
New fast and flexible sensors and actuators will need to be developed to provide plausible multimodal feedback, such as soft exoskeletons (e.g., eGloves or eBodySuits) that go beyond existing products. A high-quality multimodal feedback system should recognize and interpret the inputs from different modalities to provide multimodal outputs for human multisensory processing. Intelligent adaptive sensors are required for these purposes. Human psychophysical parameters provide the requirements for the actuators and multimodal interfaces. Psychophysical thresholds (e.g., tactile acuity or just noticeable level of difference in sound frequency) define the necessary information for developing and designing interfaces with realistic and compelling multisensory feedback that also entail tactile and kinesthetic signals. Latency sensitive haptic and visual codec applications will be required in addition to the feedback design. As a concrete scenario, novel robotic hand–arm systems will need to be developed in TaHiL that utilize haptic feedback to enable new control designs and methods for digitally transmitted remote physical manipulation and learning. The manipulation aspect of human-oriented feedback information is comparable to the approach in MP3 to avoid unnecessary information in audio. The haptic feedback will also be used to connect to smart wearables to immerse the human user in virtual or augmented reality. To achieve fast sensing and actuating for multimodal feedback, a new generation of electronics will need to be developed. It has to provide real-time operation, while consuming very low energy. Moreover, to facilitate the natural interaction between the human and the CPS, which are both equipped with high number of sensors and actuators, the electronic hardware has to be ultracompact, bendable, and stretchable. The transceiver dimensions are strongly determined by the antenna size, which massively decreases with increasing frequency. In this regard, by using very high frequencies, around 100 GHz, the antenna area can be decreased by several orders of magnitude allowing on-chip integration. To minimize the average power consumption of these wireless transceivers to below 1 mW and to allow real-time operation, aggressive duty cycling with record circuit reaction time in the nanosecond regime, allowing very efficient sleep and wake-up modes, would be necessary. The huge amount of sensor data has to be preprocessed adaptively and locally to guarantee low latency. For this, a mobile high-performance body computing platform will need to be designed with significantly reduced power consumption (20 TOPS/W). To allow mechanical flexibility, the chips will need to be thinned and integrated on stretchable substrates. Many of these new technologies are being used and developed in CeTI (see Chapters 10 and 12 for details).
1.2.3 Objective 3: Human–machine networks
Develop completely softwarized network solutions for wireless and wired communication that provide low latency, resilience, and security to enable human–machine co-operation.
To realize the vision of TaHiL, it is necessary to develop novel communication and network solutions for the body area, local area, and wide area networks that go beyond the mere relaying of information in the wireless and wired domain. New concepts of softwarization, including software defined networks, network function virtualization, and information-centric networking, will need to be exploited with respect to low latency and learning capabilities [8]. The latter is needed to allow the network to learn about different multimodal information flows (i.e., audio, video, and haptic) to adapt the network capabilities (e.g., network slicing). Moreover, the intelligent network will need to provide the means for learning from human behavior. For instance, when the connection to the human gets lost, the intelligent network should predict action courses during the period of interrupted communication (e.g., in a mobile edge cloud). Furthermore, the human's high-level but abstract mental models of behavior as well as low-level sensorimotor programs that make it possible for the human to cope with delays, ambiguities, and disruptions of the technical world will need to be better understood when designing and implementing communication networks. Such knowledge is a prerequisite for the design of efficient human–technology feedback strategies on all levels and provides an end-user-oriented direction for solving technical conflicts between latency, bandwidth, and resilience. Resolving this challenge has the potential to tremendously accelerate the progress in this field (for details see Chapters 6 and 11 for research activities on these topics that are currently being pursued in CeTI for details).
1.2.4 Objective 4: Human–machine learning
Provide an integrated framework that leverages the effects of continuous, mutual adaptive learning between humans and machines. Tune explanation facilities towards the demands and objectives of the human user. Assess boundary conditions and benefits for skill acquisition and training.
To achieve this objective, it is necessary to develop a portfolio of models, methods, and tools for automated human-inspired computing, to provide the Human-in-the-Closed-Loop interactive system with explanation and learning assistive equipment. In our view, the main innovation will be a new model-based approach that incorporates recent neuroscientific achievements on nonlinear dynamic models for human decision-making. To ensure smooth closed-loop human–machine interaction, it is important to enable human and machine to learn to predict and support each other's actions online and thereby bringing interaction to a new era of immersed closed-loop human–machine cooperation and coaugmentation. On the one hand, human actions can be augmented by machines; on the other hand, machines can learn from human behavior to represent and generate expert knowledge with various methods. The software systems for TaHiL applications have to provide self-explanation techniques to help end-users to understand the correct function and the rationale of decisions that are taken by machines. Towards an integrated framework, model-based explanations for machine behaviors, explanations for machine-learning results, and their appropriate visualization to facilitate human understanding will need to be developed. To this end, the TaHiL framework (see Figs. 1.13 and 9.1) can be applied to different domains of competency acquisition to explore the interplay of individual and situational factors to identify the conditions under which best benefits can be obtained for human–machine learning and coadaptation (see Chapter 8 for details).
1.2.5 Objective 5: Human–machine computation
Deliver a secure and scalable computing infrastructure that enables intuitive haptic interaction and automatically adapts to changes in task contexts and world models.
The softwarization of the TI leads to highly immersive software and services. The software infrastructure not only needs to perform fast, but also needs to be highly resilient to failures and attacks. Thus safety, security, privacy, and scalability are prerequisites for the tactile computing infrastructure serving TaHiL applications. The software for these applications, being naturally embedded into the physical environments, will need to rely on plausible world and context models that can also adapt to changes (see Chapter 13 for details).
1.2.6 Objective 6: Human–machine communication
Provide novel coding and compression methods, such as haptic codecs, compressed sensing, and network coding that take into consideration human factors to enable a combined control and communication system.
The TaHiL approach will produce a massive amount of sensor and actuator information to be exchanged among multiple communication and control nodes to facilitate human–machine cohabitation. Currently, control systems employ almost exclusively wired communication, because wireless solutions suffer from the traditional hard trade-off between latency, resilience, and throughput. However, the TaHiL applications need to rely on wireless communication to enable higher degrees of freedom for humans and machines. The massive amounts of sensor and actuator data that will be produced thus calls for a novel compression method—haptic codecs [9], which is akin to source compression for audio and video, but this time for all possible types of haptic information. Beside this end-to-end approach, new distributed approaches, using network coding and compressed sensing will soften the aforementioned trade-off, allowing for more flexible optimization strategies. Both approaches need to be tailored to the software-defined networks and network virtualization solutions that have to be developed. Furthermore, the amount of sensing and feedback data will need to be reduced significantly through learning strategies and optimizations of the communication network and control loop. The amount of traffic needed for closed-loop human–machine interactions in the TI will need to be reduced by at least one, but potentially up to two, orders of magnitude compared to the current standard. In addition, haptic communication requires qualitative and quantitative assessments of the user's quality-of-experience. In contrast to time-consuming subjective tests, reliable automated quality metrics for the evaluation of human-in-the-loop systems with haptic feedback will need to be developed. A range of research activities is currently being undertaken in CeTI to resolve these challenges (see Chapters 5, 6, and 11 for details).
1.3 A synergistic research program
To achieve the six research objectives and tackle the challenges presented above, we suggest a research program to facilitate synergies between the different fields for transdisciplinary research (see Fig. 1.15). Specifically, we identified five Talent Pools (TPs) that together build the foundation for the research on TaHiL. They span research activities on (i) human perception and action, novel sensors and actuators, core networking technologies, flexible electronics for on-body computation, as well as (v) computation and computing infrastructure for the TI.
Fig. 1.15 A synergistic research program involving 12 interlinked research groups for the research on TaHiL.
Through close collaborations, researchers from the disciplines of the TPs engage in developing Key Technologies and Methods (K), particularly with regards to (i) codecs for tactile and haptic communication, secure, intelligent networks, novel user interfaces, and human–machine mutual learning techniques. Outcomes from basic research conducted by the research groups at the TP-level and novel technologies developed by research groups at the K-level can, in turn, be integrated and tested in research groups at the Use Cases (U)-level (see Fig. 1.15). We have identified three broad use cases (U) involving application in (i) medicine (see left panel of Fig. 1.16), industry (see middle panel of Fig. 1.16), and the Internet of Skills (see right panel of Fig. 1.16). Other application domains can be further integrated into the research program. However, the three domains we focus on here are complete in the sense that requirements for future research in the field of TaHiL for other applications can be derived from them.
Fig. 1.16 (left) The vision for medicine and health. (middle) The vision for industry 4.0. (right) The vision for internet of skills.
An instantiation of this suggested research program is CeTI, which has been established at TUD since 2019. The structure of the research program (Fig. 1.15) illustrates the dependencies and synergistic interplays among the various groups of experts to progressively enable solutions from highly specialized to increasingly interdisciplinary teams. Outcomes of the solutions are then evaluated in different application domains, through which new assumptions and requirements are derived and transmitted back to the lower levels of the research structure for further research and technological refinements. Here we also use this structure to organize the chapters in this book. Key themes and issues indicated in the first three parts of the book and tackled by each of the twelve U, K, and TP research groups are presented in detail in corresponding chapters. Furthermore, topics on standardization, digital trace data library and communicating technological developments to the public are also treated in separate chapters in the last part of the book.
1.4 Research outreaches and societal impacts
Other than scientific and technological impacts,