Senseable City Lab :.:: Massachusetts Institute of Technology

This paper might be a pre-copy-editing or a post-print author-produced .pdf of an article accepted for publication. For

the definitive publisher-authenticated version, please refer directly to publishing house’s archive system

SENSEABLE CITY LAB

Humanizing Digital RealityDesign Modelling Symposium Paris 2017

Editors : Klaas De Rycke, Christoph Gengnagel, Olivier Baverel, Jane Burry,Caitlin Mueller, Minh Man Nguyen, Philippe Rahm, Mette Ramsgaard Thomsen.

Klaas De Rycke • Christoph GengnagelOlivier Baverel • Jane BurryCaitlin Mueller • Minh Man NguyenPhilippe Rahm • Mette Ramsgaard Thomsen (Editors)

Humanizing Digital RealityDesign Modelling Symposium Paris 2017

123

Editors

Klaas De RyckeEcole Nationale Supérieure d’Architecturede Versailles

VersaillesFrance

Christoph GengnagelUniversity of the ArtsBerlinGermany

Olivier BaverelEcole des Ponts ParisTechChamps sur MarneFrance

Jane BurrySwinburne University of TechnologyMelbourneAustralia

Caitlin MuellerSchool of Architecture + PlanningMassachusetts Institute of TechnologyCambridge, MAUSA

Minh Man NguyenENSA Paris-MalaquaisParisFrance

Philippe RahmPhilippe Rahm architectsParisFrance

Mette Ramsgaard ThomsenThe Royal Danish Academy of Fine Arts,Schools of Architecture, Design andConservation

CopenhagenDenmark

ISBN 978-981-10-6610-8 ISBN 978-981-10-6611-5 (eBook)https://doi.org/10.1007/978-981-10-6611-5

Library of Congress Control Number: 2017952525

© Springer Nature Singapore Pte Ltd. 2018

The papers are published in the form submitted by the authors, after revision by the Scientific Committee andre-layout by the Editors. The Editors cannot accept any responsibility for the content of the papers.This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of thematerial is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,broadcasting, reproduction on microfilms or in any other physical way, and transmission or informationstorage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology nowknown or hereafter developed.The use of general descriptive names, registered names, trademarks, service marks, etc. in this publicationdoes not imply, even in the absence of a specific statement, that such names are exempt from the relevantprotective laws and regulations and therefore free for general use.The publisher, the authors and the editors are safe to assume that the advice and information in this book arebelieved to be true and accurate at the date of publication. Neither the publisher nor the authors or the editorsgive a warranty, express or implied, with respect to the material contained herein or for any errors oromissions that may have been made. The publisher remains neutral with regard to jurisdictional claims inpublished maps and institutional affiliations.

Printed on acid-free paper

This Springer imprint is published by Springer NatureThe registered company is Springer Nature Singapore Pte Ltd.The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721, Singapore

Contents

Part I Material Practice

Computational Material Cultures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5Achim Menges

Part II Structural Innovation

Make Complex Structures Affordable . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17Jean-François Caron and Olivier Baverel

Part III Data Farming

Data Morphogenesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31Kasper Jordaens

Mutually Assured Construction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41Usman Haque

Seven Short Reflections on Cities, Data, Economy and Politics. . . . . . . . . . . 47Tomas Diez

Part IV Data Shaping Cities

What Big Data Tell Us About Trees and the Sky in the Cities . . . . . . . . . . . 59Fábio Duarte and Carlo Ratti

Urban Sensing: Toward a New Form of Collective Consciousness? . . . . . . . . 63Antoine Picon

Part V Thermodynamic Practice

On the Nature of Thermodynamic Models in Architecture and Climate . . . . . 85Nadir Abdessemed

The Tool(s) Versus The Toolkit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93Chris Mackey and Mostapha Sadeghipour Roudsari

Capturing the Dynamics of Air in Design . . . . . . . . . . . . . . . . . . . . . . . . . 103Jane Burry

Part VI Scientific Contributions

Stone Morphologies: Erosion-Based Digital Fabrication Through Event-Driven Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113

Steve De Micoli, Katja Rinderspacher, and Achim Menges

Designing Grid Structures Using Asymptotic Curve Networks . . . . . . . . . . . 125Eike Schling, Denis Hitrec, and Rainer Barthel

The Grand Hall of the Elbphilharmonie Hamburg: Development ofParametric and Digital Fabrication Tools in Architectural and AcousticalDesign . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141

Benjamin S. Koren

Bloomberg Ramp: Collaborative Workflows, Sharing Data and DesignLogics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153

Jonathan Rabagliati, Jeroen Janssen, Edoardo Tibuzzi, Federico DePaoli, Paul Casson, and Richard Maddock

Nature-Based Hybrid Computational Geometry System for OptimizingComponent Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167

Danil Nagy, Dale Zhao, and David Benjamin

Enabling Inference in Performance-Driven Design Exploration . . . . . . . . . . . 177Zack Xuereb Conti and Sawako Kaijima

Aspects of Sound as Design Driver: Parametric Design of an AcousticCeiling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189

Moritz Rumpf, Markus Schein, Johannes Kuhnen, and ManfredGrohmann

La Seine Musicale. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201Sylvain Usai and Hanno Stehling

The Design Implications of Form-Finding with Dynamic Topologies . . . . . . . 211Seiichi Suzuki and Jan Knippers

City Gaming and Participation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225Areti Markopoulou, Marco Ingrassia, Angelos Chronis, and AurelRichard

The Potential of Shape Memory Alloys in Deployable Systems—A Designand Experimental Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237

Philippe Hannequart, Michael Peigney, Jean-François Caron, OlivierBaverel, and Emmanuel Viglino

XVIII Contents

A Multi-scalar Approach for the Modelling and Fabrication of Free-FormGlue-Laminated Timber Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247

Tom Svilans, Paul Poinet, Martin Tamke, and Mette RamsgaardThomsen

Assessment of RANS Turbulence Models in Urban Environments: CFDSimulation of Airflow Around Idealized High-Rise Morphologies . . . . . . . . . 259

Farshid Kardan, Olivier Baverel, and Fernando Porté Agel

Automated Generation of Knit Patterns for Non-developable Surfaces . . . . . . 271Mariana Popscu, Matthias Rippmann, Tom Van Mele, and PhilippeBlock

Enlisting Clustering and Graph-Traversal Methods for Cutting Pattern andNet Topology Design in Pneumatic Hybrids . . . . . . . . . . . . . . . . . . . . . . . . 285

Phil Ayres, Petras Vestartas, and Mette Ramsgaard Thomsen

Simulation and Real-Time Design Tools for an Emergent Design Strategyin Architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295

Ole Klingemann

Robotic Fabrication Techniques for Material of Unknown Geometry . . . . . . . 311Philipp Eversmann

Locally Varied Auxetic Structures for Doubly-Curved Shapes. . . . . . . . . . . . 323Jan Friedrich, Sven Pfeiffer, and Christoph Gengnagel

This Room Is Too Dark and the Shape Is Too Long: QuantifyingArchitectural Design to Predict Successful Spaces . . . . . . . . . . . . . . . . . . . . 337

Carlo Bailey, Nicole Phelan, Ann Cosgrove, and Daniel Davis

Modelling and Representing Climatic Data in the Tropics: A Web BasedPilot Project for Colombia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 349

Roland Hudson and Rodrigo Velsaco

Thermally Informed Bending: Relating Curvature to Heat GenerationThrough Infrared Sensing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 361

Dorit Aviv and Eric Teitelbaum

Localised and Learnt Applications of Machine Learning for RoboticIncremental Sheet Forming. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 373

Mateusz Zwierzycki, Paul Nicholas, and Mette Ramsgaard Thomsen

Energy Efficient Design for 3D Printed Earth Architecture . . . . . . . . . . . . . . 383Alexandre Dubor, Edouard Cabay, and Angelos Chronis

Contents XIX

Ice Formwork for Ultra-High Performance Concrete: Simulation of IceMelting Deformations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 395

Vasily Sitnikov

Toward a Spatial Model for Outdoor Thermal Comfort . . . . . . . . . . . . . . . . 407Evan Greenberg, Anna Mavrogianni, and Sean Hanna

Survey-Based Simulation of User Satisfaction for Generative Design inArchitecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 417

Lorenzo Villaggi, James Stoddart, Danil Nagy, and David Benjamin

Navigating the Intangible Spatial-Data-Driven Design Modelling inArchitecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431

Jens Pedersen, Ryan Hughes, and Corneel Cannaerts

Tailoring the Bending Behaviour of Material Patterns for the Induction ofDouble Curvature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 441

Riccardo La Magna and Jan Knippers

Design of Space Truss Based Insulating Walls for Robotic Fabrication inConcrete . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 453

Romain Duballet, Olivier Baverel, and Justin Dirrenberger

Collaborative Models for Design Computation and Form Finding—NewWorkflows in Versioning Design Processes . . . . . . . . . . . . . . . . . . . . . . . . 463

Jonas Runberger and Julian Lienhard

Modelling Workflow Data, Collaboration and Dynamic Modelling Practice . . . 479Kåre Stokholm Poulsgaard and Kenn Clausen

Equilibrium-Aware Shape Design for Concrete Printing . . . . . . . . . . . . . . . . 493Shajay Bhooshan, Tom van Mele, and Philippe Block

Monolithic Earthen Shells Digital Fabrication: Hybrid Workflow . . . . . . . . . 509Maite Bravo and Stephanie Chaltiel

Application of Machine Learning Within the Integrative Design andFabrication of Robotic Rod Bending Processes . . . . . . . . . . . . . . . . . . . . . . 523

Maria Smigielska

Data and Design: Using Knowledge Generation/Visual Analytic Paradigmsto Understand Mobile Social Media in Urban Design . . . . . . . . . . . . . . . . . 537

Eric Sauda, Ginette Wessel, and Alireza Karduni

Simulating Pedestrian Movement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 547Renee Puusepp, Taavi Lõoke, Damiano Cerrone, and Kristjan Männigo

Structural Patterning of Surfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 559Renaud Danhaive and Caitlin Mueller

XX Contents

Free-Form Wooden Structures: Parametric Optimization of Double-CurvedLattice Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 573

Klaas De Rycke, Louis Bergis, and Ewa Jankowska-Kus

The Mere Exposure Effect in Architecture . . . . . . . . . . . . . . . . . . . . . . . . . 589Michael-Paul “Jack” James, Christopher Beorkrem, and JeffersonEllinger

[Kak-Tos]: A Tool for Optimizing Conceptual Mass Design and Orientationfor Rainwater Harvesting Facades . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 603

Christopher Beorkrem and Ashley Damiano

Redundancy and Resilience in Reciprocal Systems . . . . . . . . . . . . . . . . . . . 613Joseph Benedetti, Pierre André, Audrey Aquaronne, and Olivier Baverel

Negotiating Sound Performance and Advanced Manufacturing in ComplexArchitectural Design Modelling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 627

Pantea Alambeigi, Canhui Chen, and Jane Burry

Pneu and Shell: Constructing Free-Form Concrete Shells with a PneumaticFormwork . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 639

Maxie Schneider

A Computational Approach to Methodologies of Landscape Design . . . . . . . 657Ignacio López Busón, Mary Polites, Miguel Vidal Calvet, and Han Yu

Design Tools and Workflows for Braided Structures . . . . . . . . . . . . . . . . . . 671Petras Vestartas, Mary Katherine Heinrich, Mateusz Zwierzycki, DavidAndres Leon, Ashkan Cheheltan, Riccardo La Magna, and Phil Ayres

Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 683

Contents XXI

What Big Data Tell Us About Treesand the Sky in the Cities

Fábio Duarte(&) and Carlo Ratti

Massachusetts Institute of Technology, Cambridge, USA{fduarte,ratti}@mit.edu

Since Google Street View (GSV) was launched in 2007, its cars have been collectingmillions of photographs in hundreds of cities around the world. In New York City alone,there are about 100,000 sampling points, with six photographs captured in each of them,totaling 600,000 images. In London, this number reaches 1 million images. The GSV fleetnow also includes bicycles, trolleys (for indoor spaces), snowmobiles, and “trekkers” (forareas inaccessible by other modes). Using the images to fly over the Grand Canyon, visithistoric landmarks in Egypt, discover national parks in Uganda, or circulate through thestreets ofMoscow, although great experiences, explore only themost immediate and visualaspects of the images. Such an overwhelming abundance of images becomes much moreinteresting when we consider them as a rich source of urban information.

Researchers in the fields of computer sciences and artificial intelligence have beenapplying computer vision and machine learning techniques to interpret GSV images.Very few of them move beyond the technical aspects of deciphering these images toexplore novel ways to understand the urban environment. The few examples includethe detection and counting of pedestrians (Yin et al. 2015) or the inferring of landmarksin cities (Lander et al. 2017). Still, most of this research is either based on small subsetsof GSV data or presents a combination of techniques in which the participation ofhumans is required:

At the Senseable City Lab, we have been using computer vision and machinelearning techniques to analyze full datasets of GSV images in order to understand urbanfeatures in ways that would take too long or be financially prohibitive for most citiesusing human-based or other technological methods. We started by looking at the treesand to the sky. Exposure to greenery and natural light is essential to human well-being,

Fig. 1 Computer vision process

© Springer Nature Singapore Pte Ltd. 2018K. De Rycke et al., Humanizing Digital Reality,https://doi.org/10.1007/978-981-10-6611-5_6

outdoor comfort, and climate mitigation. Therefore, quantifying green areas and lightexposure in different parts of the city will inform better urban design as well asenvironmental and public health policies. By using GSV data with computer visiontechniques, we demonstrate the value of bringing big data to the human level, to thetangible aspects of urban life.Usually, street trees are quantified and characterized using field surveys or othertechnologies such as high spatial resolution remote sensing. These techniques dependon intensive manual labor, specialized knowledge, and ad hoc data acquisition.Although satellite imagery analysis gives accurate quantification and characterizationof green areas in cities, the technology has two critical caveats for urban dwellers:firstly, it looks at the city from above, not from a person’s perspective. Satellite imagerydoes not show greenery at the street level, which is the most active space in the city andwhere people see and feel the urban environment. Secondly, larger green areas arehighlighted in detriment to the relatively sparse street greenery. However, visits toparks and urban forests do not happen frequently and the benefits of these areas are feltat a large scale, whereas street trees are part of citizens’ daily experience and haveimmediate positive effects on people’s lives. We are not dismissing such techniques,but finding ways to take advantage of the huge amount of standardized visual datafreely available of hundreds of cities to propose a human-centric and comparableassessment of street greenery.

Using large GSV datasets composed of hundreds of thousands of images per city, Li et al.(2015) and Seiferling et al. (2017) calculated the percentage of green vegetation in streets,using computer vision techniques to detect green pixels in each image and subtractgeometric shapes. With a few computational steps, what is left from this subtraction isgreenery. Since theGSVdata acquisition procedure is standard, thesemethods allow us to

Fig. 2 Treepedia in Frankfurt

60 F. Duarte and C. Ratti

calculate street greenery in dozens of cities around theworld and to compare them—usingwhat we called the green view index.1

By avoiding the pitfalls of creating “algorithmic sorting of places” (Shapiro 2017),which automates the attribution of social values onto visual aspects of an image, theanalysis of large visual datasets with the same computer vision techniques acrossdifferent cities and countries has the power to become a civic tool, by which citizenscan compare street greenery in different cities and neighborhoods and demand adequatemeasures from public authorities.

A recent work (Li et al. 2017) has applied similar techniques to measure the skyview factor in cities. The sky view factor is usually understood as “the ratio betweenradiation received by a planar surface and that from the entire hemispheric radiatingenvironment” (Svensson 2004: 203), varying from 0 to 1. In cities, it can used toquantify the degree of sky visibility within urban canyons, by which one can infer theexposure to natural light in each site, for instance. A common technique to measure thesky view factor is to capture fisheye images with special cameras. Again, as thistechnique is time consuming—and therefore financially prohibitive for most cities—even when it is done, it usually covers only part of the city. We have been usingcomputer vision algorithms to analyze GSV panorama images in order to optimize theprocess, cover the entire city, and make such analysis more accessible.

Besides using sky view factor as an indicator of local environmental conditions, atthe Senseable City Lab we are exploring using it in order to optimize urban infras-tructure. One example is optimizing energy-saving programs in public areas. Citieshave been converting their traditional street lights into LED technology, which con-sumes less energy and save cities millions of dollars per year—the 26 million streetlights in the USA consume more than $2 billion in energy, and the greenhouse gasemissions they generate is comparable to 2.6 million cars. However, in most cities,even in those converting streetlights to LED, unless lampposts are equipped withphotosensors, all streetlights turn on automatically at the same time, in some casesvarying daily according to the astronomical sunset. Applying computer vision tech-niques to analyze dozens of thousands of GSV images, we can determine the sky viewfactor at each data point and match them with the nearby streetlights. By accounting forbuildings and trees blocking the adequate amount of lighting required in each point ofthe city, it would be as if we had hyperlocal sunsets close to each streetlight and coulddetermine the optimal time to turn on the lights, which would save energy and moneyto cities at an aggregate level. Using this highly granular information, we could opti-mize existing infrastructures without adding another layer of devices, but rather byusing data which is already available.2

The underlying research question is how not to take data at face value but insteadby the intrinsic information they hold about how cities work and how citizens live inthe urban environment. A GSV image is more than simply a combined photograph ifyou analyze it with the appropriate tools. In both cases discussed here—street greenery

1 Treepedia project is available at http://senseable.mit.edu/treepedia.2 We are grateful to Ricardo Álvarez and Xiaojiang Li for some of the ideas discussed here; and toLenna Johnsen for revising the paper.

What Big Data Tell Us About Trees … 61

and sky view factor—it is possible to imagine that soon such large amount of visualdata will be collected more frequently and in many more cities. Furthermore, with moresensors deployed in urban infrastructure, embedded in personal mobile devices, andsoon in driverless cars, we can foresee all this data available in real-time maps, whichwill help to design actuations at the local level as well as enable the creation ofworldwide urban dashboards that would show multiple cities in a comparative way.Making sense of the sizeable quantities of data that is already generated in and aboutour cities will be key to creating innovative approaches to urban design, planning, andmanagement.

References

Lander, C., Wiehr, F., Herbig, N., Krüger, A., Löchtefeld, M.: Inferring landmarks for pedestriannavigation from mobile eye-tracking data and Google Street View. In: Proceedings of the2017 CHI Conference Extended Abstracts on Human Factors in Computing Systems—CHIEA’17 (2017)

Li, X., Zhang, C., Li, W., Ricard, R., Meng, Q., Zhang, W.: Assessing street-level urban greeneryusing Google Street View and a modified green view index. Urban For. Urban Greening 14(3), 675–685 (2015)

Li, X., Ratti, C., Seiferling, I.: Mapping urban landscapes along streets using google street view.In: Patterson, M. (ed.) Advances in Cartography and GIScience, Lecture Notes inGeoinformation and Cartography. DOI: 10.1007/978-3-319-57336-6_24 (2017)

Seiferling, I., Naikc, N., Ratti, C., Proulx, R.: Green streets—quantifying and mapping urbantrees with street-level imagery and computer vision. Landscape Urban Plann. 165, 93–101(2017)

Shapiro, A.: Street-level: Google Street View’s abstraction by datafication. New Media Soc.146144481668729 (2017)

Svensson, M.K.: Sky view factor analysis—implications for urban air temperature differences.Meteorol. Appl. 11(3), 201–211 (2004)

Yin, L., Cheng, Q., Wang, Z., Shao, Z.: ‘Big data’ for pedestrian volume: exploring the use ofGoogle Street View images for pedestrian counts. Appl. Geogr. 63, 337–345 (2015)

62 F. Duarte and C. Ratti