Hi
| Tianyu Su |
City | Technology | Design
Projects
| Data | + | Urban Technology |
LivingLine
City Science Group,
MIT Media Lab
2019-2020
LivingLine Shanghai is a version of the MIT CityScope platform for urban analysis, efficient resource utilization, and spatial programming. I am leading the Wi-Fi data analysis and helped build the tangible, augmented reality platform used to visualize complex urban relationships, simulate the impact of multiple urban interventions, and support decision-making in a dynamic, iterative, evidence-based process.
# Data Analysis, Machine Learning, Localization, Augmented Reality, 3D Design and Modeling
Presentation
International Conference on Urban Studies (Cambridge, UK, 2020)
| Data | + | Choice Modeling |
Tailor to Fit
MIT Office of Sustainability
MIT Urban Mobility Lab
2019 - 2020
Increasing commuting time and costs due to congestions and fixed working hours have long been a problem for faculty and employee. Also, The institutions are troubled by the difficulties of providing enough heavily subsidized parking. However, based on the MIT Commuting Survey, individual model choices and their responses towards the same TDM like AccessMIT vary a lot, which allows us to think about specific targeting programs for different subgroups.
With the belief in mind that institute-wide TDM program may not fit everyone, I start to look at the possibility of “tailored and targeted” commute-related recommendations and programs, by looking at a case study project related to Commute Behavior Preferences Reshaping and Mode Shift Potentials at MIT.
# Data Analysis, Machine Learning, Web Mapping
| Computer Vision |
Tasty City
Advances in Computer Vision, EECS, MIT
2019
Photo Geolocation using Street and Restaurant Data.
In computer vision, the photo geolocation problem has been usually approached at a global scale or regional
scale. In this project, we derive knowledge from urban studies and present a classification based method looking into neighborhood scale photo geotagging.
We subdivide Boston Chinatown into multiple geographic cells and train a deep network using 20K google street view images labeled with local restaurant density and street speed limit. We show that this multitask model could output a probability distribution over a couple of cells in the neighborhood.
# Computer Vision, Deep Learning, Mapping
| Deep Learning | + | Data |
MIT Parking Prediction
Deep Learning for Transportation, DUSP, MIT
2019
In this project, we focus on predicting the number of daily MIT gated parking coming from each census tract in Parking Year 2017-2018 (Sep. 16, 2016 - Sep. 15, 2018), as well as the overall number of daily parking of all parking permit holders. We trained a dual-stage
attention-based recurrent neural network ( DA-RNN) model with the campus parking history and exogenous variables such as weather, holidays, transit health, and use the model to predict the daily number of parking.
# Deep Learning, Spatial Analytics, Mapping
| Data |
Spatial Innovation
Department of Urban Studies and Planning, MIT
2018-2019
Within the context of the late 2000s, wide diffusion of innovation hubs emerges. Spatial Innovation project studied the distribution of innovation hubs (incubators, accelerators & co-working spaces) in San Francisco and key spatial factors influencing them by extracting spatial data and deploying logistic regressions.
# Spatial Analytics, Data Visualization, Web Mapping
Oral presentation
16th International Conference on Computers in Urban Planning and Urban Management (CUPUM 2019)
| Data | + |Interactive Design|
Inno-
Immigration
Department of Urban Studies and Planning, MIT
2019
Immigrants play an increasingly pivotal role in the U.S. economy. While there is contentious political debate regarding if immigrants are taking away jobs from the locals, there is little discussion about how immigrants could contribute to job generation and economic growth.
In this study, we explored the immigrants' entrepreneurship activities in the last decades by looking at their geographical and industrial distribution. And we argue that a startup visa could create a pathway to draw foreign-born founders to the US and improve its competitiveness when other countries are launching more start-ups.
# Data Analysis and Visualization, Web Development, Interactive Design
| Data |
Innovation
Field Lab
Ash Center,
Harvard Kennedy School
2019
In this collaboration between City Form Lab and Harvard Kennedy School's Innovation Field Lab New York project, we featured housing challenges in 10 cities around New York Area by compiling GIS data and developing a series of maps.
We also explored root causes behind "problem properties" by mapping vacant properties, fire incidents, call for services, code violations, as well as historical Red Lining boundaries.
# Spatial Analysis, Mapping
Poster presentation
City+2019: The International Conference for Ph.D. Students and Early Career researchers on Urban Studies
| Data |
Airbnb & Ghost Hotels
Department of Urban Studies and Planning, MIT
2019
From 2009, short-term rentals like Airbnb growing statewide, which poses serious questions to local communities and city decision-makers. In this research, we dug into four research questions by spatial clustering (DBSCAN) and propensity score matching:
1) Where is Airbnb activity located? And how is it changing? 2) Who makes money from Airbnb? And how much are they making? 3) How much housing has Airbnb removed from the market? and 4) What’s the monetary incentives for the Airbnb hosts to lease their short-term rental listing as entire homes, rather than private rooms?
# Spatial Analysis, Machine Learning, Panel Regression, Web Mapping
| Data | + | Design |
SHOE
as a FIELD
New Balance Innovation Studio
2018-2019
Shoe as a Field project experimented a holistic data-driven approach to redesign the midsole. Outsole and insole pressure data was applied to shape the form of midsole units and the distributions of them.
# Data Analysis and Visualization, 3D Design and Modeling, Computational Design
| Media | + | Design |
Eco-LIVE
Arts @ MIT
2019
Eco-live attempts to address the issue of unequal distribution of education resources in different school districts and communities, improve learning experience, and create a sharing education platform that promotes cross-district education equity, through a combination of available augmented reality technology, simulation tools, and open source online learning materials.
# Augmented Reality, 3D Design and Modeling, Product Design, Entrepreneurship
First Place
MIT Connect Arts, Community, and Computing Challenge
Honorable Mention
Design Intelligence Award | Digital Interaction
| Data | + | Urbanism |
Panyu
Interactive
Tsinghua University
2017-2018
To improve the undesired urban environment around Shanghai Film Center, Panyu Interactive use scientific research methods and data analysis in understanding the relationship between human behaviors and the urban environment. Rather than a brand new urban design, we proposed a fully programmed open space system transplanting the social activities onto streets to make a safer, playful, and interactive community with minimum urban interventions.
# Urban Design, Spatial Analytics
Chapter 6
Handbook on Planning Support Science
| Urbanism |
Biosynthesis
Harmony
Tsinghua University
2017-2018
This project illustrates a harmonious biosynthetic neighborhood, regarding ecology as an approach of urban regeneration. As the Yongsan Animal Protection Center was plugged in the negative edging area between the previous community Haebangchon and the USAG Yongsan, the entire area is activated with biological diversity and sustainable social effects.
# Urban Design, Architecture Design and Modeling, Computational Design
About
Tianyu Su
A mixed-methods researcher and entrepreneur at the intersection of urban, human, and technology, with a hope to shape better cities by a deeper understanding of society and human behaviors.
Doctor of Design (Urban and Design Technology) | Harvard University
Master in City Planning | MIT
Master of Architecture, Bachelor of Architecture | Tsinghua University
Quantitative Design Researcher | Harvard University
Researcher| MIT Media Lab
Data Analysis and Visualization
R, Python, SQL
Spatial Analytics
ArcGIS, QGIS, PostGIS, SQL, CARTO, Mapbox
Machine Learning
PyTorch, scikit-learn
Web Development
HTML, CSS, JavaScript, D3
Augmented Reality
Unity 3D, ARCore, Vuforia
3D Design and Modeling
Rhino, Grasshopper, SketchUp
Rendering
V-Ray, Lumion, Keyshot
Graphic
Photoshop, Illustrator, InDesign
2D Prototyping
Sketch, AutoCAD
