Table Of Content

Construction and Building Automation This book is intended to be used as a textbook in undergraduate civil engineering and construction courses to introduce cutting edge mechanical, electrical, and computer science topics that are needed for civil and construction engineers to collaborate in inter-disciplinary automation projects. Part I introduces the basics of hardware and software technologies that are needed for implementing automation in buildings and construction. The content begins with the fundamental concepts and uses practical examples to bring out the benefits of automation through case studies that are easy to understand. No other book uniformly treats the subject of automation within the context of buildings and construction activities. While the technology needed for these two application domains are similar, the unifying principles are not well recognized. This book will bring out the fundamental principles that could form the basis of application to these two domains. For example, it will become clear that sensors, actuators, and controllers, along with smart control strategies could be used for automating tasks within buildings and on construction sites. Part II of the book will introduce key advances in the areas of machine learning and artificial intelligence that are significant for the intelligent control of buildings and construction equipment. Control algorithms and techniques for data analytics are explained in a form that is appropriate for non-computer science students. Each chapter contains several hands-on exercises meant to apply the principles that are covered. These include numerical problems as well as design and analysis examples. This new textbook: • Introduces hardware and software needed for automating engineering tasks • Presents examples of applications in the control of building systems • Illustrates of the use of automation for improving construction processes • Provides a lucid introduction to advanced computing concepts, machine learning, artificial intelligence, and control algorithms to construction and engineering students It is sure to be essential reading for a growing number of courses in smart construction, building automation, robotics, intelligent buildings, and construction 4.0. Supplementary material including answers to exercises in the book will be provided on the author’s website: https://bennyraphael.com/book2022/ Benny Raphael is a professor in the Department of Civil Engineering at IIT Madras, India. “A comprehensive treaty on a growing area of knowledge.” Professor Abhijit Mukherjee, Curtin University, Australia Construction and Building Automation From Concepts to Implementation Benny Raphael Cover image: © Benny Raphael First published 2023 by Routledge 4 Park Square, Milton Park, Abingdon, Oxon OX14 4RN and by Routledge 605 Third Avenue, New York, NY 10158 Routledge is an imprint of the Taylor & Francis Group, an informa business © 2023 Benny Raphael The right of Benny Raphael to be identified as author of this work has been asserted in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging-in-Publication Data Names: Raphael, B. (Benny), author. Title: Construction and building automation : from concepts to implementation / Benny Raphael. Description: Abingdon, Oxon; New York, NY : Routledge, [2022] | Includes bibliographical references and index. | Summary: “This book is intended to be used as a textbook in undergraduate civil engineering and construction courses to introduce cutting edge mechanical, electrical and computer science topics that are needed for civil and construction engineers to collaborate in inter-disciplinary automation projects” — Provided by publisher. Identifiers: LCCN 2022002858 (print) | LCCN 2022002859 (ebook) | ISBN 9780367761271 (hbk) | ISBN 9780367761103 (pbk) | ISBN 9781003165620 (ebk) Subjects: LCSH: Building—Data processing. | Building—Mathematics. Classification: LCC TH153 .R28 2022 (print) | LCC TH153 (ebook) | DDC 690.0285—dc23/eng/20220413 LC record available at https://lccn.loc.gov/2022002858 LC ebook record available at https://lccn.loc.gov/2022002859 ISBN: 978-0-367-76127-1 (hbk) ISBN: 978-0-367-76110-3 (pbk) ISBN: 978-1-003-16562-0 (ebk) DOI: 10.1201/9781003165620 Typeset in Times New Roman by Apex CoVantage, LLC Contents Foreword vi Preface viii Acknowledgments x PART I 1 1 Introduction to automation 3 2 Hardware for automation 15 3 Software for automation 51 4 Application: building automation 69 5 Application: construction automation 111 PART II 135 6 Introduction to machine learning 137 7 Basic mathematics for machine learning 153 8 Regression 179 9 Classification task 198 10 Inductive learning – decision trees and random forests 239 11 Unsupervised learning algorithms 248 Epilogue 264 Index 265 Foreword In Building and Construction Automation, Professor Raphael delivers the fundamentals of this rapidly growing and evolving domain clearly, cohesively and cogently. Senior under- graduates and masters students in civil, architectural, construction, environmental, and mechanical engineering programs will find this text a valuable introduction on which they can build with both experience gained in the field and additional courses in building science, construction robotics, infrastructure computer vision, smart structures, modularization, circular engineering in the built environment, generative building design, and data analytics. Students in mechatronics, and in electrical and computer engineering will use this text to ground and provide a built environment context for their work in intelligent control systems, robotics, machine learning, computer vision, and mixed-reality technologies. Professor Raphael reduces the overwhelming complexity in this fluid domain without oversimplifying. For example, automation is reduced to hardware and software, and the domain is reduced to the act and the result of building. Civil engineers who have taken a cir- cuits and instrumentation course will know enough to appreciate and absorb the explanation of hardware elements such as power sources, sensors, actuators, controllers, communication devices, and robotics. Having taken an introductory computing course, they will further absorb the importance of wireless and wired communications, the central role of protocols, and the need for control systems, including the PID (proportional integral-derivative) control algorithm that remains ubiquitous in practice, despite the existence of highly sophisti- cated alternatives. A humanist theme permeates the text. A smart building is defined as one in which information and communication systems work together with building systems to maximize the satisfaction of its occupants. While it is recognized that stakeholders have differing perspec- tives (to an owner, the building is an investment that must pay off), the role of the occupant is foremost. Construction automation and robotics present similar moral dilemmas. They may improve quality and project performance in terms of cost and schedule; they may even improve safety. Will they take away jobs, or are they an inevitable evolution that will result in more skilled and highly paid jobs? Professor Raphael addresses such concerns head-on in the text, and by doing so keeps us focused on the dignity of the builders and the wellbeing of society. Humanity recognizes the imperative of sustainability to the world’s wellbeing. Buildings require vast amounts of abiotic materials and energy to construct. For example, they demand 50% of global steel production. They are responsible for 40% of the solid waste produced by our societies. Operating them requires up to 40% of the energy produced by our econo- mies. Mass and energy must be manipulated in extremely large quantities in the act and result of building; this is unavoidable. As we seek to reduce this footprint through circular Foreword vii engineering in the built environment, building construction and automation will play a key role. For example, modularization is identified in the text as conducive of automation and as a critical element of design for automation, design for reuse, and design for disassem- bly. Deep renovations and retrofitting of existing buildings to reduce their operating energy depend on the principles of building automation explained in this text. In summary, this text supports what should be a required pair of courses in most civil, construction, and architectural engineering programs. It is a substantial contribution to the knowledge base. Prof. Carl Haas, University of Waterloo, Canada (Former President of the International Association for Automation and Robotics in Construction – IAARC) Preface The book is primarily aimed at civil engineers, architects, and project managers. They may not be that much interested in the proofs of theorems and other mathematical details. They want to know what techniques are available and where they can be used. Hence mathematical proofs and derivations are omitted, except for some important concepts in which the derivation or proof is essential to understand the fundamental principles involved. The book is not written for computer scientists or mathematicians. They are likely to be disappointed by this book because it does not have the mathematical rigor that they expect. However, the book aims to teach fundamental principles rather than software tools. It is not advisable to train engineers to use toolkits as black boxes without understanding the details of the underlying algorithms. They need to be clear about what are the assumptions involved, under what circumstances the techniques can be applied, and the limitations of the techniques. This requires a thorough understanding of the mathematics and the fundamental principles. This is especially true about topics related to machine learning. The primary message from this book is that not all methods are suitable for all tasks. It requires considerable knowledge and experience to formulate the problem correctly and to apply the right techniques to solve them. Since this book is meant to be an undergraduate or postgraduate textbook, exhaustive literature reviews are not provided. The objective is mainly to introduce basic concepts. Occasionally, references are provided to research papers, mostly based on the author’s research. These are meant for students who wish to go deeper into application examples that illustrate the concepts described in the book. Special emphasis is placed on solving numerical problems. Working out numerical problems helps in understanding fundamental mathematical concepts. Some of the examples and exercises might be too simple for students who have a strong mathematical background. However, in my experience as a teacher, I find that there is a wide spectrum of students in every class. Some students may not be strong in mathematics, but they have brilliant ideas and make good engineers. Such students need more exercises and worked-out solutions to understand the concepts that seem obvi- ous to more mathematically inclined students. The book is written to cater to the require- ments of both categories of students. Students with strong mathematical background can simply skip the examples that are too elementary. Other students get a chance to refresh their knowledge of mathematics without having to search for these topics in mathematics textbooks. Introductory chapters are written without much mathematical content, for example, Chapter 6. These chapters could be used to give an overview of concepts to a non- mathematical audience. The mathematical details are provided in subsequent chapters. Engineers, architects, and project managers have a considerable amount of special- ized domain knowledge that is essential for implementing automation in buildings and Preface ix construction projects. However, in most colleges, they do not learn automation technologies or machine learning. This type of inter-disciplinary knowledge is essential for working in projects involving automation. They need to collaborate effectively with electrical and mechanical engineers, and computer scientists. Only then we will be able to see appropriate use of automation technologies in building and construction industry. The topics covered in this book evolved over many years through teaching experience in these subjects at the National University of Singapore (NUS) and IIT Madras. In addition to the formal courses taught at these institutes, the author conducted a number of workshops under the umbrella of the International Association for Automation and Robot- ics in Construction (IAARC). Many examples and exercises included in this book were part of these workshops, which had significant hands-on content. This book could be used to replicate these workshops in other parts of the world. The numerical problems, review questions, and group discussion exercises could be used to make such workshops lively and interesting. I hope this book will help to spread knowledge of automation to civil engineers and architects, and to enable wider adoption of these technologies.