52
Engaging Public Voice in Big Data Society An on-line participatory design experiment. by Yi Hou B.L.A at Iowa State University, 2012 Submitted to the Department of Architecture in Partial Fulfillment of the Requirements for the Degree of Master of Architecture at the Massachusetts Institute of Technology February 2016 MASSACHUSETTS INSTITUTE OF TECHNOLOGY MAR 2 9 2010 LIBRARIES ARCHVES @ 2016 Yi Hou. All rights reserved. The author hereby grants to MIT permission to reproduce and to distribute publicly paper and electronic copies of this thesis document in whole or in part in any medium now known or hereafter created. Signature redacted Sig n at u re o f A utho r: .................................................................................................................. .. ..... ........ .... ... .. tmW o Architecture January 15, 2016 Signature redacted C e rtifie d b y ...............................................................................................................................-. -5 - .. - .------ --- c --- - 'U ehiko agakura Associate Professor of Design and Computation Thesis Advisor Accepted by..........................................................................................................S ignature redacted Takehiko Nagakura Chair of the Department Committee on Graduate Students

Signature redacted ignature redacted

  • Upload
    others

  • View
    14

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Signature redacted ignature redacted

Engaging Public Voice in Big Data SocietyAn on-line participatory design experiment.

by Yi Hou

B.L.A at Iowa State University, 2012

Submitted to the Department of Architecturein Partial Fulfillment of the Requirements for the Degree of

Master of Architecture at theMassachusetts Institute of Technology

February 2016

MASSACHUSETTS INSTITUTEOF TECHNOLOGY

MAR 2 9 2010

LIBRARIES

ARCHVES

@ 2016 Yi Hou. All rights reserved.

The author hereby grants to MIT permission to reproduce and to distribute publicly paper and electroniccopies of this thesis document in whole or in part in any medium now known or hereafter created.

Signature redactedSig n at u re o f A utho r: .................................................................................................................. .. ..... ........ .... ..... tmW o Architecture

January 15, 2016

Signature redactedC e rtifie d b y ...............................................................................................................................-. -5 -.. - .------ --- c --- -

'U ehiko agakuraAssociate Professor of Design and Computation

Thesis Advisor

Accepted by..........................................................................................................S ignature redactedTakehiko Nagakura

Chair of the Department Committee on Graduate Students

Page 2: Signature redacted ignature redacted
Page 3: Signature redacted ignature redacted

Thesis Committee

Thesis AdvisorTakehiko NagakuraAssociate Professcof Design and Computation

Thesis ReadersSarah WilliamsAssistant Professor of Information Technology and Urban Planning

Imdat AsAssistant Professor at University of Hartford

3

Page 4: Signature redacted ignature redacted

Engaging Public Voice in Big Data SocietyAn on-line participatory design experiment.

byYi Hou

Submitted to the Department of Architectureon January 15th, 2016

in Partial Fulfillment of the Requirements for the Degree ofMaster of Architecture

Abstract

Should the design of architecture be solely in control by architects? Should we trust the public to let them

design with us? These are the core discussion revolves around participatory design, a design approachinvolves public effort. In theory, this democratic process should results a more responsive and suitable

project for users. Although the concept of participatory design works well in theory, but there are

constrains that stopped itself from wider application. How can we collect large amount of ideas that can

directly lead to a design solution? Will the result have any advantage compare to traditional designprocess?

Building upon historic precedents, my thesis proposes an on-line participatory platform, a new computa-tional tool that allows large number of participants participate into the design process. Each user's desirecan be translated into a set of data that represents their preferences on architecture. Then, to extractcommon patterns from data pool to compute result. The final goal is to calculate an optimized design that

will suit most users'desire. Will the public like what they designed collectively? The result will spark an

interesting discussion.

Thesis Advisor : Takehiko NagakuraTitle: Associate Professor of Design and Computation

4

Page 5: Signature redacted ignature redacted

Bibliography

Negroponte, Nicholas. The Architecture Machine. Cambridge, MA: M.I.T., 1968. Print.

Negroponte, Nicholas. Soft Architecture Machines. Cambridge, MA: MIT, 1975. Print.

5

Page 6: Signature redacted ignature redacted

INTRODUCTIONBackground research 7-10

at-rn t u,

atenextacn 3

Page 7: Signature redacted ignature redacted

The stressful process of creating a user satisfied design,because architectjust don't get the idea.....

.. 2i

t ~ ~ ; 2s

I-E n,4'(\PJCY2,Lj

KA-

Sketch by Yona Fridman

o-~ 4i~!*~,-~~'2'

~-.__I

~ I H r~

~I' ~*

1~

E-~.

~ A

II cl

Yona Friedman's pictograms from Negroponte's "ComputerAided Participatory Design" in "Soft Architecture Machines"

7

* , oft

Page 8: Signature redacted ignature redacted

Early studies from Nicholas Negroponte

First image: Participant sketch out his version of the ideal house,Second image: Participant created an abstracted diagram that represents the spatial relationship between programs.Third image: Computer redrew the same diagram.Foulrth image: Computer generated plan from diagram.Fifth image: Computer reinterpreted previous diagram and generated new diagram based on the same concept.

Svth) 5maqo ompter rfqln(r1te new [plan> (M share~ the' samefl c:Onc:pt wItl dther ihs t dhi difrn crfKLquration

8

Page 9: Signature redacted ignature redacted

Traditional design approach

Owner

User s Idea

9Architect

+A&

Uesigni

C)

9ArchIitect

+

9

Participatory approach

^A,&L

Page 10: Signature redacted ignature redacted

Crowd-sourcing ProgramsPlace Pulse quantitatively measures urban perception by crowdsourcing visual surveys to users around the globe.

This program asks users to compare images collected by Google from different part of the world. By letting large

amount of users to evaluate which place do they like better, a collective big data pool is created to rank places based

on users inputs.

Which place looks livelier?

-A 4

SOe.Ot -

Less wealthyCLICKS TIND

iA-

-

:,WIMore safe

-1 T ,More wealthy

-4"

am4

Less lively ,-- _- More lively

WC.~2

10

More depre-s g Less depressing

-. ';afe_ ,

CLICS TROODRANK CITY RANK CITV

S4

55

S6

Page 11: Signature redacted ignature redacted

~=~- - - -

INTRODUCTION

Project overview 12-18

Pa ion J 8

Page 12: Signature redacted ignature redacted

Site Location: Kendal SquareKqllndl quar has - un iqu dvata on location and demographic. Its close relation with MITI has offered the potential to

become the most innovative square mile on this planet. At the same time, its location has also brought a great complexity ofdemand into its territory. A mixed group of student, tech-company employees, workers, and tourists presence at different time.Therefore, I choose to apply my participatory strategy to engage this highly diverse and dynamic neighborhood, in order to

ii' (Lt 11>1'ciffi rtFd~J

12

Page 13: Signature redacted ignature redacted

Zoning DistrictsDuring the early stage of the project, I started to introduce various constrains to

setup the computational logic. One of them is zoning regulations. The site situated

on C-1 and PUD-5 zone, which development is limited by both zoning requirments.

I

0

AC-3

C-38

C-3B MXR

C-3

MXD C2

PU 0-KS

0-2

PU 0-KS

P

C-03

~~-t

13

-3AUD-3

40

1111111P1111111V

Page 14: Signature redacted ignature redacted

C-3B PUD-5 MXR

C3B ResidenceMultifamilyLimited OfficeFAR: 3.0/4.0Height: 1 20'Minimum Area Per Units: 300 sq. ft.Maximum units per Acre: 145

PUD-5Office, Retail, HotelFAR: 2.080% builtable area, 3 story building

MXRmodifies the base residential district regulations in areas with substantial non-residential uses to encourage

residential conversion and retail and consumer services to support residential uses.

Overlay DistrictsAn overlay district is a set of zoning regulations for a defined area that are required either in addition to the base

district's regulations or in lieu of those regulations. Overlay zoning is used to protect the character of an area of

special concern or to encourage new development subject to additional controls.

14

Page 15: Signature redacted ignature redacted

15

Page 16: Signature redacted ignature redacted

......

....

""..............-.-

.

Page 17: Signature redacted ignature redacted

4

*; Ari N

_"9 SA1S0b <4

lV

4

z#q

Wlit

isj.A ~ ,-!

3,.

- 1.

AK

/

Page 18: Signature redacted ignature redacted

-AV

Page 19: Signature redacted ignature redacted

INT DTBackground r-search 7 - '

Project ovrv iew12

PARTICIPATORY PLATFORMGame development 20-26

m-age tpiatfcorm dievelopme nt 3-3Pattern extraction 37-38

Dasg frj m ata 4-141DaaVaiation 47-5)

Page 20: Signature redacted ignature redacted

LIVE 3%

STUDY 4%

Retail Demand ProjectionN poset themiC abit OkJG(J 00u .quari iee(

retail in Kendall Square. Among the existing retail

establishments in the Square, bookstores, coffeeshops, farmers markets, and food trucks were ratedas highly desirable by the community, according to

the Customer Intercept Survey done by the CommLI-nity Development Department in 2011. "work" wasthe primary purpose for respondents being in Kend-all Square. This indicates that the area has a larger

daytime population and a relatively smaller nighttime

population

20

Page 21: Signature redacted ignature redacted

List of Desired Businesses in Kendall Square

I Grocery Store

Diverse, Sit-bown Restaurants

-1 Nightl e (Bars/Music Venues/Performance Space)

7% Quick/Affordable Lunch Options

6 Convenience Store

3 Coffee Houses/Cafes

3 Electronic Goodsf~ce Supply

Other (Living, Lab, Office Space)

Apparel Stoes-

511daky Food/Liquor Store

2 Hair/Nail Salon

Publc Se_ _s y_ I y dPublic

S 0 "

F *.., -igler 0

Af fordabl e Gyms

1% Personal Services (Dry Cleaning, Doggie Day Care)

1uv yJ S b 05 conducted by t h I'ntaC

Square Initiative. This survey summarized aranking for desired programs in Kendal Squareirea.

A drugstore was the most desired retail accord-ng to the Customer Intercept Survey. The largedaytime population of the area would supply thenumber of customers needed to support such a

lsiness.[he second most-desired retail was a grocerystore, which is a key amenity to create a more;omplete neighborhood. There is insufficientdemand for a traditional supermarket due to thepresence of competitors nearby.

21

Page 22: Signature redacted ignature redacted

-F

Page 23: Signature redacted ignature redacted

P rogram Development

To further define the constrains of the game,each program block is categorized into different groups based on their size and functions.

Office

Retail space

Cafe / restaurant

COMMUNITYLandscape

Public lobby

Lecture room

MIT Classroom

Student service

Restroom

MUSEUM Exhibition space

Art gallery

U SI * *23

Page 24: Signature redacted ignature redacted

~~jI

Page 25: Signature redacted ignature redacted

Gaming Interfaceor (an Iag k o th conter of pl m bsod n huwr preference. Each program is color coded anid comes

in different modular sizes.This responsive program also communicates with user by displaying a interactive diagram at the lower right corner. This

diagram informs each user the trade offs arid impacts to the surrounding environment and budget of their proposal.The game also follows zoning regulations by taking total Gross Floor Area and height into consideration.

25

Page 26: Signature redacted ignature redacted

Model Exportr ru I be alutomict c-ily recorded an Ixrted ac 0BJ mang flcje

So they can be used for later analysis.

m - - m m - - -m m - m - - - m --

to

sa-sa u.s.. oe v.ee .... nww. r. ee9ws.

26

Page 27: Signature redacted ignature redacted

PARTICIPATORY PLATFORM

Pattern extraction 28-32

Paterrn extr ct on 37-38

DE ,G

Page 28: Signature redacted ignature redacted

1stDiagrammatic analysis penrqornation

1stRegeneration

2ndR e (r- 1, n1iza IT i "n

2ndPpgeneration

28

User's input

Page 29: Signature redacted ignature redacted

r-<.

Page 30: Signature redacted ignature redacted
Page 31: Signature redacted ignature redacted
Page 32: Signature redacted ignature redacted

.1 ~-

Page 33: Signature redacted ignature redacted

ackgrond earch7 1

PARTICIPATORY PLATFORM

Image platform development 34-36

Page 34: Signature redacted ignature redacted

p-I'I

Fl dAk

L~' IAi

'I'll'

SL)L*q. ~7r~e k.a ~Igmw , I

p Us tI~- ~

Av-~,,~

fill

rr,

:LiAbkL

Page 35: Signature redacted ignature redacted

Image Comparison Platformhe ~ 0 -I dse 'I qu sMon IdK u r O l t iding rm terials

35

Page 36: Signature redacted ignature redacted

Image Comparison PlatformI he saorm o sk user [c vote Wn ieir lavorite archi tectural massing torrn / style

)sk~ <fl$II trmstle

36

Page 37: Signature redacted ignature redacted

Background research -7-

JeCt Cverv!w r

PARTICIPATORY PLATFORM

attern exc 2-3

Pattern extraction 37-38

Desi:n crm data 40-41Data Validatn 43-50

Page 38: Signature redacted ignature redacted

iTij ~X1/TII

~--

j

17,

IAdi

- A-

l.Th* 41-

P/N/

lot-- L:41

Page 39: Signature redacted ignature redacted

IN

m

Page 40: Signature redacted ignature redacted

A

DESIGNDesign from data 40-41

Page 41: Signature redacted ignature redacted
Page 42: Signature redacted ignature redacted

iihi

9.u4

'IIi

'V4

IU

I~I

"I

Ii

~iI'IT

p

f-,flr .1

Page 43: Signature redacted ignature redacted

INTRODUi TJ0Backgr ond ref ac e 7-Project overview I2-is

PARTKC PATORY PLATFORM

Pattern e xtractcn 2 -3

-- a!e pla , f -m Idp m t 34-36

eat e X aC n 7-8

DESIGN

Data Validation 43-50

Page 44: Signature redacted ignature redacted

Data Validationn d pi )d elf evaluate the effectiveness of the platform. A data validation procedure was

required to find out whether the final design respond to user's desire.The final design was taken back to the platform for user reevaluation against other images that users has

44

Page 45: Signature redacted ignature redacted

77''

Page 46: Signature redacted ignature redacted

i~JJ

~LLI ~

I.1~

- - r ~

..I,

API

ok

Page 47: Signature redacted ignature redacted

xeig Fo

Page 48: Signature redacted ignature redacted

AI

|i_ -~

r

Page 49: Signature redacted ignature redacted

Comparison Test

49

Page 50: Signature redacted ignature redacted

Data ValidationThe s ond matidaIos th omprion test t u ;u pris resu it.

The top ranked material Ioss the comparison test. In stead, the new

50

Page 51: Signature redacted ignature redacted

Testing Failure

Do public have self-consciousness when it comes to aestheticDo public know what they really like?

The experiment was setup to compare images with isolated aesthetic properties. User were asked to

compare only one element of building at a time. But during that process, some might be distracted byother details in the image, like the building context, color of the sky or even the resolution of image. These

factors caused inconsistency in data result. Not even mentioning participant's personal instability factors,like weather, emotion or health condition might have affected their ability to make a consistent judgment.In order to eliminate these distractions, we need more thorough studies of how to extract public aesthet-

ic. Most users can not name their favorite architecture before they begin the experiment. But during the

image comparison test, everyone was able to give an answer within a relative short time. The experiment

setup helped user to organize their internal aesthetic logic, so they were able to make a final judgment.The revaluation of the platform proved the system functions as it was intended for the most part. But the

inconsistent final result also exposed the weakness of the system.

Perhaps the solutions lies in a more sophisticated technology that can interact with users to extract moreaccurate data. Imagine in the near future, the experiment will take place in ideal virtual environmentwhere distraction is limited to minimum. Users can wear VR goggle to explore their preference in a fully

immersed virtual architectural environment. Then, they would be able to make a much more accuratedecision. Such technological intervention should be able to help architects finding out the true desire of

public, to extract preferences that public don't even realize they have.

Page 52: Signature redacted ignature redacted

Today, we are one step closer to this goal. This project revolutionized participatory approach bycrowdsourcing mass data and transform them into a tangible design, at the same time, raised newchallenges to enhance data accuracy through educating public. Testing failure has a irreplaceablerole in this process of discovering new frontier, and it is our new role as an architect, to be responsi-ble and in charge of generating a more democratic and educating design process.

'4ME!3hMO

i 'N4 =gI

mm

"J, __ -, 71 T9qrIF-: RFWT"M.P..W

M