2024 MIT R&D Conference: MIT Senseable City Lab Overview

Conference Video|Duration: 10:09
November 19, 2024
  • Video details
     
    MIT Senseable City Lab: Overview
    Umberto Fugiglando
    Research Manager & Partnerships Lead, MIT Senseable City Lab
  • Interactive transcript
    Share

    UMBERTO FUGIGLANDO: How many of you-- and raise your hand-- is somehow familiar with this map? OK. For the few hands not raised, let me explain this to you. This is Trump winning.

    At a higher resolution, if you zoom in, this is somehow different. You notice that the blue dots, which are the big cities in the US, are very different from the previous map. You see that in cities, there is a prevalence of blue, which is not Trump. And we know that cities has grown-- has born more than 10,000 years ago, because people wanted to get to live together, to exchange experiences, to exchange ideas, to exchange chromosomes. And this is basically the birth of cities.

    We know today that cities behave somehow in a different way, in a special way. Well, fortunately, we can study today's cities with the help of digital technologies. This is Lisbon, seen in real-time, from the lens of data produced by the public transportation.

    We are a lab here at MIT. We are not interested in politics. We are interested in science and the science of city. We are the Senseable City Lab. We use technology to better understand what are the rules of cities, how we can converge the digital world that we live in with the physical world that we live in? And we do this all over the world, with many companies, with many cities, and with few research partners.

    So let me give you some examples of our research today. For example, something very peculiar that we find in cities is that there are a lot of environmental information that change in space and time. Take pollution, for example.

    You have very good air quality in this point of the city. You go down the block, go around the corner, and you have very bad air pollution and air quality. This is normal. It's how cities behave.

    Well, how to monitor this? We decided to build our own sensors. And instead of deploying them every meter in our cities, we equipped some cars, so that normal cars can bring their own sensors. And those are multi-purpose, multi-platform sensors.

    They are equipped with cameras. You can take pictures. You can take thermal pictures and multispectral pictures. You can have readings on temperature and humidity and, again, air quality.

    And why is this useful? This is useful, for example, to build these kind of maps. This is a way of democratizing data. You show to citizens and to companies and to municipalities, data at everybody's doorstep. What's the air quality in front of your house, in front of your kids' school?

    It's interesting that if you bring this technology to different cities or you show it to different companies, everybody finds somehow their own value. In New York, we were working with the mayor and with the municipality to look at how the underprivileged communities, the lower income communities, are disproportionately exposed to better air quality? If you go to Stockholm, they were interested in heat waves and understand how the heat bubbles in the city are created.

    I'll show you something that we're doing in Rome. And we're monitoring trees. Rome, today, has a big problem of tree health. There are small parasites that really like to eat those trees. And those very beautiful centennial trees are dying.

    So we're using the same technology, attached to cars, with few different sensors-- normal camera, multispectral camera, with some AI-- to be able to detect the tree and understand how the tree is performing photosynthesis. And we're doing this with the forest corps in Italy, the Carabinieri. So we're keeping their cars with our sensors, collecting a lot of data. And then, eventually, down the line, they're going to use satellite pictures as well to complement the view that we have from down-- from road level.

    But talking about trees, let me give you this number. 65% of insects are at risk of extinction due to climate change. This is quite dramatic. It's quite dramatic because it means that all the ecosystems that basically rule the world are at the edge of collapse. And ecosystems in cities are even more are fragile.

    And those little bugs, even if they're very annoying for you, and for me as well, they're a very good proxy of the biodiversity of a city or an environment. Counting these bugs is crucial to understand at which point our ecosystem is going up or down. Well, it turns out that all these counting are happening manually. There are people going there with a notebook, noting the bugs. There is very, very little AI on this.

    So in our lab, we said, let's try to not only recognize automatically the recent insect in this picture, but we built a platform. We built a library, an open source library, that is able to categorize the insect species. And this may sound very trivial or a computer science exercise. But then we talked with some companies-- with an engineering and architecture company-- and they were saying, well, we actually need to monitor the biodiversity in rural developments and projects. So why don't you develop something more?

    So we decided to embed this intelligence in a sensor and build, together with some experts in biology, some gardens-- some physical gardens-- with the right mix of trees and flowers to attract these insects and do the counting automatically. Those are some of the research that we do. We partner with many companies, all over the world, with many cities.

    In three of them, the one that you see in the map-- Dubai, Amsterdam, and Rio de Janeiro-- we actually build some MIT labs. So we have permanent MIT researchers there, to really create an ecosystem of innovation, together with the local stakeholders.

    In Amsterdam, for example, we're working on self-driving boats. We call them roboats because they are robotic boats. This is a long project that we are developing with them. You see here some of the true videos. If you go to Amsterdam, you can try some of them.

    And not only they are autonomous for people, but with the same kind of technology, you can use-- you can address other city needs, like trash collection, for example. The challenge here was really to design the technology for enabling a urban boat to go around, to design it, and to understand what are the urban uses for this. And we've done everything, from the hardware to the software to the urban cases.

    If you go to Rio de Janeiro, for example, we're interested in this kind of environments, favelas. The shanty towns-- a town that has grown without rules. And that's the result.

    Well, many of the planning techniques here are failing. Amazon doesn't deliver here. Why? Because there are no street address. There is no cadaster.

    So we're using new technologies. We were the first ones to create 3D maps, with LiDAR, of these environments. And this is the base for a digital twin of a favela, that eventually enables us to give property titles to people; understand what's the airflow; understand what's the risk of TB; and plan for surgical interventions, instead of just demolishing everything like many politicians are planning for.

    So those are some of the research that we are doing here at the lab. We are lucky enough to do, together with ILP, our big conference, our annual flagship conference of the lab, here on campus, happening tomorrow. It's called "Being Physical" because this year, we're really going to explore how the physical proximity of people in cities are fostering for a more resilient environment and for a more resilient society.

    It's going to be in-person only. Many people ask me, is there a remote connection? No, it's being physical.

    So it's going to be a physical conference here on the MIT Samberg Conference Center. It's going to be on two days. So first day, tomorrow-- tomorrow afternoon-- we're going to have some panels, with some experts on these topics. And then in two days, on Thursday, in our lab, we're going to have a demo day, where we're going to show some of these technology. We're going to do some deep dives into some of these projects.

    And for ILP members, there is complementary access. So just use this QR code. Use the ForumILP2024 code and you're able to register. It's, again, free for ILP.

    And we're really looking forward to welcoming you tomorrow, at Samberg here, and, in general, to develop a conversation with our researchers, for maybe starting some research topics, some research explorations, with you guys. Thank you so much.

    [APPLAUSE]

  • Video details
     
    MIT Senseable City Lab: Overview
    Umberto Fugiglando
    Research Manager & Partnerships Lead, MIT Senseable City Lab