Science lab spotlight

Navigating our cities

Using technology to build smarter cities

9178 hubcab01
The HubCab project revealed that only an average of 10 taxis cover a third of Manhattan’s streets daily.
Courtesy of the Senseable City Lab

With new advancements in technology and the abundance of data, we can better understand the interactions between people and their urban environments. As a result, improvements in urban planning can pave the way for more efficient and environmentally cleaner cities. Researchers at the MIT Senseable City Lab aim to predict and study these improvements from a critical point of view. As conducting research to learn about people’s habits in their urban environment requires members of the lab to consider many diverse viewpoints, the Senseable City Lab is made up of a multidisciplinary team of designers, engineers, computer scientists, biologists, and social scientists. With this diversity of researchers comes a diversity of technologies being utilized in the lab. “Reflecting the diversity of the lab, and the urban issues, we use big data analysis, machine learning techniques, but also robotics and design,” says the director of the lab, Professor Carlo Ratti. 

The team collects and analyzes two types of data: opportunistic data and sensor data. Opportunistic data is always being produced from social media logs, cell phone calls, and WiFi connections. On the other hand, sensor data is collected through devices developed by the lab for the specific problem at hand. They’ve applied their sensors to prevalent environmental problems in urban areas, such as deploying sensors on garbage trucks to measure the air quality of cities. Additionally, they’ve integrated their sensors on small robots to collect waste water for analysis of the sewage system. 

The research conducted at the Senseable City Lab has produced many exciting findings. For one, using the public data of the 150 million taxi trips that take place in New York, they found that only ten taxis cover a third of Manhattan’s streets daily. “Thus, by attaching inexpensive sensors to crowd-sourced urban vehicles, we can capture hyper-local measurements across a large portion of a city,” says Ratti. The lab has also utilized 50 million geo-tagged Weibo (a social media platform) check-ins as a measurement of social activity in specific locations. By correlating this data with the daily pollutant records for 251 cities, the lab found that urban activities are uniquely affected by air pollution. “For instance, we found a greater effect on locals’ activities [compared] to visitors and on locals’ leisure activities compared to work activities,” says Ratti. 

In addition to these projects, Ratti’s lab is currently working on two projects entitled “AI Station” and “Tasty Data.” The “AI Station” is a collaboration with the Société Nationale des Chemins de Fer (SNCF) Gares & Connexions at the Gare de Lyon train station in Paris. This project studies the digital footprint that people leave in spaces to better understand passenger behavior, with the end goal of improving their experience at train stations. The data collected consisted of Wi-Fi information from the 100 million passengers that pass through the Gare de Lyon train station annually. The Wi-Fi data was used as a digital footprint to aid in understanding the behavior of passengers in train stations. 

The “Tasty Data” project utilizes restaurant data to make predictions about local socioeconomic factors. Restaurant data included ratings, cuisine types, and more, and was provided by Dianping, the Chinese version of Yelp. The lab built a model that takes in this restaurant data and predicts factors like daytime and nighttime populations, company presence, and spending amounts. These predictions were calculated across 9 cities in China to compare their socioeconomic attributes.

In light of the increasing availability of globally representative data, Ratti is hopeful about the future of the lab; “As digital technologies are pervasive, as well as urbanization, we see the lab working more and more in problems that present global patterns — and this can help us to address urban problems that happen everywhere.”