Sensible City
Methodology Overview
Course: MIT 11.S951 Senseable City: Data and Analytics
Instructor: Carlo Ratti, Fabio Duarte, Paolo Santi
Individual Work
Date: Feb.-Apr., 2022
Since their emergence around 10,000 years ago, cities have evolved into the most magnificent and consequential artifacts of human culture. Today, Big Data and new computational tools are empowering us to study the forces that shape cities quantitatively for the first time in history.
This course helps uncover the laws of urban life through the exploration of new methods borrowed from experimental physics and computational sciences. The resulting equations and models hold the keys to our cities — and to our common urban future.
​
Specific questions this course will try to address are:
-What brings us together or tears us apart in cities?
-How do we move, socialize and behave in cities?
-How does data science help us uncover these hidden patterns, and inform the design of better cities?
KEYWORDS:
Big Data, Quantitative Analysis, Urban Environment, Data Science, Future Development
OBJECTIVES
-An understanding of the spatial scale of urban analysis and data
-An ability to interpret big data through technical approaches
-An ability to critically evaluate the results and findings of data analysis
-Knowledge of artificial intelligence applied to urban sciences
-A critical approach to mobility and urban design to avoid segregation
01 Visual AI
Fundamental notions related to the spatial and temporal patterns of signals/sensed quantities collected in urban environments will be presented. Notions of spatial and temporal granularity will be introduced, as well as digital techniques for visual data acquisition and analysis to understand the urban environment. During this lesson, the Helsingborg case study will be presented: the Swedish city is collaborating with the Senseable City Lab to improve public safety management.
View Full Notes
Google Colab
- Instruction Notes View
- Demo_0__GSV_download.ipynb View
- Demo_1__Classifier.ipynb View
- Demo_2__Autoencoder.ipynb View
- Demo_3__Object_Detection.ipynb View
- Demo_4__Image_Segmentation_solution.ipynb View
02 Mobility
The lesson will start from the paths connecting cities, analyzing new mobility possibilities including on-demand mobility systems, ride sharing opportunities, and their impact on traffic and parking infrastructure. The methodology part will cover the fundamental tensions of Paris urban movements, how we can uncover the laws of our cities through new methods, the use of artificial intelligence as a tool both to collect data and analyze it, and the creation of simulations for multiple applications. The data are the result of a collaboration with Paris’ transport agency, RATP, with the goal of better understanding and designing the city.
03 Social Segregation
Urban data can help us understand many different aspects and dynamics of cities. In this class, we will analyze in particular the social behaviors determined by the urban system and morphology, with a case study on inclusion and segregation. The methodology part will focus on Paris as a case study.
04 City Scanner
From urban dynamics scaling to indoor spaces, sensing and collecting data can tell us a lot about people’s behaviors. During the class will be presented different research projects conducted at MIT Senseable City Lab during the last ten years. During this class we will present the City Scanner project, which is an ongoing research initiative with deployments in several cities around the world. An overview of this research will be followed by a hands-on workshop to acquaint students with the data it has produced. In particular, the engineering challenges of developing sensors will be discussed as it relates to the acquisition of urban data.
