Urban phenomena through digital traces

Fábio Duarte - MIT Senseable City Lab, USA

Abstract

As layers of networks and digital information blanket urban space, new approaches to the study of the built environment are emerging. The way we describe and understand cities is being radically transformed—as are the tools we use to design them. In this presentation, I will explore how we can use digital tools to understand urban phenomena in novel ways. In Desirable Streets, I will explore how we can quantify what makes some street more desirable than others by analyzing hundreds of thousands of actual pedestrians trips in Boston, based on data collected via self-tracking apps. In Favela 4D I will show how we’re combining community involvement with handheld laser scanning technologies and methods to mathematically understand the complex urban environment of informal settlements in Rio de Janeiro.

Machine Learning and Event Detection for Urban Public Health

Daniel B. Neill - Machine Learning for Good Laboratory, New York University, USA

Abstract

This talk will present new machine learning algorithms which can assist cities in monitoring and improving population health. First, we describe new approaches for targeting health interventions to address the opioid crisis. Our methods provide situational awareness of emerging patterns of overdoses, opioid use disorder, and risk behaviors at the geographic, subpopulation, and individual levels, as well as identifying networks of physicians and dispensaries engaged in unsafe opioid prescribing behaviors. Second, we describe a new event detection approach that provides public health practitioners with a "safety net" to detect newly emerging disease outbreaks and other previously unknown but potentially relevant patterns, using free-text chief complaint data from hospital emergency departments. Together, these methods will provide city public health agencies with a suite of tools and methods to improve the overall quality of population health and reduce health disparities.

Do higher order connections matter? Measuring urban and regional industry agglomeration patterns

Neave O' Clery - UCL, UK

Abstract

This talk will discuss a number of recent papers on the theme of measuring urban and regional industry agglomeration patterns. In particular, we will focus on the role of high order inter-industry connections, as encoded in the topology of industry networks or ’spaces', typically ignored in most common measures of industry agglomeration. Using data for the US, Ireland, Colombia and the UK, we show that these connections can play a key role in predicting a range of quantities such as local industry entry and employment growth.

Making cities better with human-centric urban data science

Michael Szell - IT University of Copenhagen, Denmark

Abstract

Making our cities better and sustainable is key to solving the climate and urban transport crises. To this end, urban data science offers new tools to quantify societal problems in cities and to propose human-centric solutions to policy makers. In this talk I outline our recent and ongoing efforts towards that end, focusing on urban transport and vulnerable road users. I discuss inequalities between transport modes through mobility space and collision threat distributions, and automated methods based on network science to generate and merge bicycle networks and to identify their gaps, for different urban development stages. All research comes with concrete policy recommendations that significantly increase urban livability, public health, and decarbonization.

What-if scenarios for urban regeneration

Vittorio Loreto - Director of the SONY Computer Science Lab - Paris

Abstract

Modern cities are at the centre of a passionate debate about their future. The ongoing pandemic has driven many inhabitants out of cities and into safer, less crowded areas and poses a real challenge to authorities, both in terms of urban planning and policy decisions. There is a need to rethink the role of cities, adapting their infrastructure and dynamics to the 'new normal'. It is then of paramount importance to tackle the challenges that urban areas face going beyond pure optimisation schemes, keeping a transformative eye. New tools are thus needed, allowing for a realistic forecast of how a change in the current conditions will affect and modify the future scenario. In this talk, I will present the so-called "what-if" machine, a recently proposed platform that provides users with tools to assess the status of urban and inter-urban spaces and conceive new solutions and new scenarios. I will describe a few examples, ranging from modelling at the coarse-grained level of urban socio-economic variables to the microscopic level of mobility to new urban scenarios like the 15-minutes city.