Hi there!

I’m Romy - a designer, researcher and recent Harvard grad, dedicated to rethinking humane technology and experimenting with any new tool that comes my way. 



Find out more:
︎ Information
︎ Linkedin
︎ Resume















Featured            UX Design            Code            WorldBuilding



COMPUTATION AS TRUST

MIT Architecture | 2018
Instructor:
Takehiko Nagakura, Axel Kilian
Duration: 2 months
Tools used: Python, CAD

︎ Problem:
Apartment rental in cities is getting more and more expensive. Viable housing options for single families and/or couples and young adults are becoming more rare and coliving might not be the most desirable situation for all.

︎ Solution: Match-making occupants based on their opposing lifestyles, providing them with a larger apartment when needed without having them share spaces.




Problem




Apartment rental in cities is getting more and more expensive. Viable housing options for single families and/or couples and young adults are becoming more rare and coliving might not be the most desirable situation for all.


How might we design a coliving model that is both affordableand preserves the privacy of single families, couples and young adults?




Approach




01
Define

Online search
Survey 1

02
Research

User Survey 2
Chosen Personas
03
Design

Code: scoring
Application in bldgs
UI Interface

04
Speculation

What does this model mean for the future of living?




Research

Phase 1: Co-Living preferences & problems



Phase 2: Survey

Middle and Low income single families and/or couples and young adults  interested in renting a large apartment where they can feel well-off and are able to fulfill their needs without the need to share spaces with others.
45 people in Boston answered the questionnaire that had 40 questions. Through this survey, I wanted to understand to the smallest detail the lifestyles of these people living in Boston. Questions asked ranged from working schedule, to usual social hours, to hours of practicing a hobby.  The goal was to translate the findings to two sets of quantitative data: preferences in co-living conditions, and occupancy patterns.




Phase 3: User Survey


_01_

50% preferred living alone

_02_

70% didn’t mind only having access to a few rooms, if the others are in use by a roommate

_03_

80% would prefer having a larger living space only occasionally

_04_

75% would prefer paying for rent in spaces used if accessibility is limited on other days.


Phase 4: Chosen Personas

Sibylle 26 Female, Architect (only uses her living space a few nights per week)
Olga 35 Female, Dance Instructor (uses her living space for classes)
Renee & Fouad 33,36 Female Male, Bankers (living early morning, and early afternoon)
Salem 40 Male, Consultant (stays late every night and uses living space Sunday night)
Joseph 35 Male, Engineer (has gatherings 4 nights a week)
Katerina 28 Female, Musician (late sleeper, uses living room till 3 am)
Eli 26 Male, Student (loves cooking and having people over during lunch time)
Yara & Sasha 28,29 Female (Y is away on weekends, S uses the living on afternoons)
William 27 Male, Consultant (travels during weekends)


The 9 people’s lifestyles were translated into 0s, and 1s of occupancy. I then proceeded to visualize their living pattern in a typical apartment: the wireframe is a 0 (unoccupied) and the filled box is a 1 (occupied).









Solution


Main idea


What if we could pay rent according to usage of spaces? Or had a larger apartment only when we have guests and a smaller when we are by ourselves?

This Data-Driven Co-Living model locates people closer to their most compatible neighbors (most different from their lifestyle). That way common spaces are shared with exclusive access to each whenever the space is theirs.



Code:  Finding compatibility score & Visualization

To get compatibility scores, the columns of living rooms and kitchen (the bedrooms were kept as totally private) were summed with other personas, and when the result was 0, the compatibility score increased, when it was 1, it also increased, and when it was 2, it decreased by twice as much.

 




Application in existing buildings

To understand the extent to which shared spaces can occur, I studied 3 buildings in Boston. The first was Peabody Terrace, the second was The Hayden Building and the third was Douglas Park.  Each had different possibilities according to typologies that would end up creating different spaces for users.




User Interface

Register lifestyle and reserve an apartment


initial sketches



How does residential scheduling affect behavior?



Future work - Computation as Trust? 




With an interest in the social aspect of this research, I decided to reflect on the sharing economy in living. What would happen if keys were replaced by computation?


Can computation eliminate barriers of trust in housing if spaces were physically inaccessible whenever used by another party?

Can computation replace keys if we optimize for individual usage of space? What kind of world would that look like?