Senseable Design

Carlo Ratti on mining big data for health

Article image

The director of the Carlo Ratti Associati design and innovation practice and director of the MIT Senseable City Lab shares his vision of how digital technologies and healthcare analytics can inform a more proactive approach to improving health.

Your work focuses on the use of new technologies to improve life in cities. How can it be used to improve health and well-being?

Health and well-being cannot solely be reduced to a numeric factor, but data sourced from digital technology can help us to become more aware of something. If I learn about my behavior, then I can improve it; it is a way to create healthy feedback loops, both at the individual and the social scale. I think the quantified self-approach is a positive example of how gathering data about ourselves can both encourage a more active lifestyle and encourage social dynamics.

The smart city concept is often criticized for its elitist, top-down and de-humanizing approach. How can we avoid technology’s potentially negative effects on human psychology, well-being and on the health of society as a whole?

Yes, there have been criticisms of this concept, and I often agree with them. That is why I prefer the term senseable city which I believe takes a more humanistic approach: it better encapsulates the possible social benefits coming from embedding Internet-of-Things technologies into our urban spaces, as opposed to technology per se. Senseable implies both the ability to sense through digital technologies, and the more human quality of being ‘sensible’, of keeping people and their desires at the center. Ultimately it is an issue of approach — top-down versus bottom-up.

The Mario prototype could provide health data to enrich healthcare analytics and respond to outbreaks rather than being reactive
The Mario prototype waste-sampler being lowered into the sewers of Cambridge, Massachusetts, as part of the Underworlds project. It is one of two probes nicknamed Mario and Luigi after the pipe-walking Super Mario Bros. Photo courtesy Carlo Ratti.

At MIT Senseable City Lab, you are working on various projects related to healthcare analytics. Can you tell us about a few of them?

One of the most recent examples is: Weibo Smog. We used data from the Chinese equivalent of Twitter, Weibo, to understand the sensitivity of different areas of China to air pollution, according to their socioeconomic profiles. We hope that by studying and reporting on the differing ways in which air pollution actually affects Chinese citizens and their health, policy officials and others can then be in a better position to address this concern.

“The goal is to use this information to help health clinics, hospitals, health care providers and the general population to prepare and proactively respond to outbreaks rather than being reactive.”

Your project Biobot is using sewer robots to track a major public health crisis: the opioid epidemic in the US. How does it work and what kind of progress has been made so far?

Biobot Analytics is the company that came out of a project we ran in 2016 called Underworlds. The premise of Underworlds is that you can tell a lot about a person by sampling their microbiome—essentially their germs. As this information is flushed down the toilet, it lives on in our sewage: a kind of collective gut. The project proposed to tap into this wealth of information and was designed to detect and characterize viruses, bacteria and chemicals in a city’s sewage. We built small robots to source samples from urban sewers in order to detect fluctuations in urban health parameters. The goal is to use this information to help health clinics, hospitals, health care providers and the general population to prepare and proactively respond to outbreaks rather than being reactive. Until now, we have sampled in Boston, Kuwait City, and Seoul. Biobot, which is continuing the mission of this project, is preparing the next deployments. Currently, it is focused on using these analyses to estimate opioid consumption in towns across the United States.

Have any of the other initiatives seen a measurable impact on health? Can you give us some examples?
We focus more on the urban environment and its quantification. However, individual medical records are becoming increasingly available, and they will allow us to close the loop, so to speak. I believe that the combination of individual data on environmental exposure, behavior and health will be a game changer in medicine. We started working with this type of data on a project with GE Health called Health InfoScape, which aims to use data – in particular, the electronic (anonymized) medical records from across the United States – to uncover human health patterns and discover statistical relationships between regions, types of spaces and illnesses through healthcare analytics.

Research consistently reveals how minorities, low-income or poor communities are more exposed to health risks than whiter and wealthier segments of the population. How can technology contribute to social and environmental justice?
I think the best way to expose these biases is through data. That will help us to correct them.

Healthcare analytics can inform a more proactive approach to improving health
Office interiors of the Agnelli Foundation in Torino, Italy. Photo courtesy Carlo Ratti

What do you think are the greatest challenges to improving health through design?
Health tends to align many interests —individual, collective, corporate, etc.—so I do not see major challenges here. But there are still some technical challenges. How do we integrate data at the individual level?

An example might be our renovation of the Agnelli Foundation HQ in the heart of Torino. In collaboration with tech company Siemens, we equipped the century-old building with hundreds of digital sensors that monitor different sets of data, including occupancy, temperature, carbon dioxide level and the status of meeting rooms. Using a mobile app, users can access the monitoring platform that enables actions such as interacting with other people, checking in, booking meeting rooms and regulating environmental settings, all with an unprecedented degree of personalization.

For instance, once a building occupant sets her preferred temperature and illumination settings, the BMS responds accordingly, adjusting the levels of lighting, heating and cooling. As the fan coil units situated in the false ceilings are activated by human presence, the system can potentially follow occupants as they move around the building, just like an “environmental bubble.” When an occupant leaves, the room returns to “standby mode” and conserves energy – as a computer would do with a screen saver.

In this sense, the challenge here is how to manage huge and anonymous data sets while keeping the focus on individuals’ specific needs.

Main image: The Luigi prototype for the Underworlds project. Courtesy Carlo Ratti.

Weibo Smog