On April 27th 2022, the Emerging Leaders of URBAN AI gathered for a lecture by Carlo Ratti, on the topic of Senseable Cities. In addition to his role as professor at MIT, Carlo Ratti is also a partner at Carlo Ratti Associati, a globally active innovation and design firm. In his lecture, Ratti spoke about a wide range of projects that have enjoyed his collaboration over the past decades. In this blog post, Emerging Leaders Céleste Richard and Lars Kouwenhoven will first introduce the reader to topics and projects featured in the lecture, and furthermore reflect on its contents.
Systematizing the Senseable City
Throughout time, people have had an interest in the future. Specifically, many have attempted to predict the future. These predictions remain fiction more often than not: there are infinitely many potential futures, and the likelihood of predicting exactly the correct one is nihil. Ratti quotes Karl Popper, noting:
“The future is open. It is not predetermined and thus cannot be predicted — except by accident. The possibilities that lie in the future are infinite”
Rather than predicting the future, Carlo Ratti attempts to explore the future, through design and science. When focusing on the topic of cities, such explorations are equivalent to thinking about the different ways in which cities may evolve in the future. The term design as used by Ratti refers to a broad definition that spans disciplines such as mathematics, complex science and machine learning.
One central theme in Carlo Ratti’s explorations through design is the convergence of the digital and physical world. The internet, once existing as an abstract concept, is now entering the physical space, with the advance of techniques such as the Internet of Things, deploying an array of sensors throughout cities. The data points produced by such sensors can then be used to analyze and further understand processes taking place in cities.
What type of explorations has such data enabled? One early example from 2005 uses real-time mobile phone data from Rome, Italy, to analyze and graphically communicate urban processes and patterns. While such early projects focused on just one, or a handful, of locations, recent work has enabled exploration at scale. The universal visitation law of human mobility, a paper co-authored by Ratti and Schläpfer (more info below), describes patterns in mobility that could be revealed for the first time thanks to a large body of mobility data. Such large-scale studies were unthinkable until just a couple of decades ago.
Other forms of data may also help reduce inefficiencies in cities. Thanks to publicly available taxi pick-up and drop-off data in New York City, the MIT Senseable City Lab was able to describe how many taxi drops could have potentially been combined, owing to the fact that they took place from similar pick-up, to similar drop-off locations, around the same time of day. Their research found that they could satisfy the mobility demand of Manhattan’s ride-hail industry with just 60% of taxis. Thus Ratti took an existing form of technology, and analyzed how its efficiency could be improved. However, rather than leaving the technology as it is, he has also worked on expanding existing technology, such as using existing fleets of vehicles and equipped them with sensors. As a result, the capability of the vehicles is expanded, without having changed its initial functionality — this project will be further elaborated in the next section.Not only it is about exploring how emerging technology can be used to sense a city, but also how emerging technology can change the city itself.
Analyzing the Senseable City
The Senseable City, from its name, aims at sensing cities to capture its essence and understand its functioning. From a big data collection, to an in-depth analysis, the senseable approach aims at using the results to transform different dimensions of the city. The goal is to find efficient and smart ways to live in and with the city. There is an interest in bridging people, technologies and cities. To foster future responsive cities.
Data offers the possibility to discover the city, its activity and its inhabitants from a different perspective. Data can shed light on how people live in the city and inform on their needs. What could be improved? What could be tackled more efficiently? As mentioned previously, the New York City taxi project aimed at reducing city congestion. This challenge, when tackled successfully, has a positive impact on the user experience. As a matter of fact, reducing cars by developing a car pooling system, results in reducing traffic and -noise- pollution, therefore offering a better experience of the city.
Scanning the city can also be coupled with already-existing city elements, aiming at using and optimizing the current infrastructure to extract its characteristics. The project City Scanner works in this direction, collecting real-time data from sensors placed on garbage trucks in the city of Cambridge (US). Garbage trucks are therefore transformed into sensing platforms, measuring air quality, facade temperature, air temperature, humidity and road quality. After processing, analyzing and visualizing the data collected, the project aims at helping people understand the spaces they interact and live every day, offering an open city — accessible to all.
In addition to efficiently tackling city challenges, data can also help to tend towards a tailored city, based on its inhabitants’ needs and comfort. While looking at the USA system, cities have a strong car-centric approach resulting in a lot of parking spaces and traffic. By analyzing people’s flows and activities, we can manage the use of space in cities, saving parking space and aim at on-demand mobility.
These senseable solutions indeed aim at developing a tailored city, but they also aim at fostering an automated way of experiencing the city. The project Roboat in Amsterdam (NL) is the perfect example of embracing and implementing technology for a pioneering city. The city of Amsterdam is made up of many canals, offering a new vision on transportation. In cities, we are used to using tramways, metro, bikes, cars to navigate from A to B, but what about using the city’s particular infrastructure in order to reach a destination? By developing a self-driving boat, the Senseable City Lab tackles once again the automated and on-demand approach in cities. Inhabitants can use these smart mobility solutions in order to reach their destination at any moment of the day, reducing traffic in the streets and contributing to a zero emission transportation mode. Moreover, while navigating on the canals, real-time data is collected, optimizing the use of this self-driving boat and giving insights on the canal’s infrastructure.
By collecting data to sense the city we live in, the Senseable City aims at fostering a data-driven approach for efficient, smart and responsive future cities. By combining uses, by understanding and linking the digital and physical layers of this living organism, the Senseable City aims at offering tailored cities that respond to its inhabitants’ needs.
Reflecting on the Senseable City
One key consideration in the deployment of many of the processes described, is what the potential privacy implications of such interventions may be. Is there such a thing as too much sensing? A potential avenue is to keep the intensity of sensing constant, but improve the communication related to it. A company like Numina, which enables cities to monitor and sense their streets, makes their sensors prominent, rather than hide them. In a way, many sensors have existed in an unequal relationship: while they are capable of perceiving every move, the individual is not even aware of their presence. Changing this relationship may be a first step in the direction of more productive and fair sensor usage.
If one were to put privacy implications aside, further questions could arise regarding the degree to which present data availability informs, or limits, potential deployments. Projects by initiatives such as the senseable city lab are inherently limited, both in fiscal as well as in physical manners. While the City of Boston can afford to equip its garbage trucks with heat sensors, this is less likely to be the case in many locations throughout the Global South. What type of issues are we overlooking as a result of this?
A key assumption underlying much of the research related to senseable cities, is the idea that social patterns can be digitally sensed. While it certainly is true that people’s movement patterns can be recorded and sensed, and certain social habits subsequently can be derived, sensors cannot infer why individuals partake in these movements. Certainly, one can attempt to infer this information, but it is hard to get to a ground truth.
Deploying sensors in cities on already existing elements is well thought out, as it creates a smart, resilient and sustainable approach in collecting data on cities and its inhabitants. This thinking process could be exhaustively tackled to improve quality of life for people, with a user-centric perspective based on live dataset collection. Such datasets, coupled with proper expertise can help city planners, municipalities and other stakeholders in enhancing and managing livability.
Finally, we would argue that one of the most important tasks of a project that uses sensors to sense human behavior, is to subsequently communicate the outcomes of the project, as well as focus on how it can tangibly improve the quality of life of these individuals. These projects could be a first step in the transition from enterprise-owned to personally owned data, empowering the many.
By Céleste Richard (Junior Planner at ARUP Sustainable Cities & Transport and Contributor at URBAN AI) and Lars Kouwenhoven (Urban Tech Hub Coordinator at Cornell University and Contributor at URBAN AI)
