Este artículo está también disponible en / This post is also available in: Spanish
One of the most important quotes in any business is, “if you can’t measure it, you can’t manage it”. Today, more than ever, this quote is also true for cities where the rise in use of connected devices in urban space has immensely increased the amount of data generated – 90% of data available in 2016 was created in the two years prior[1]. This growth will continue to be exponential as technology becomes cheaper and more readily available, presenting a unique opportunity for cities to use this data to improve decision-making processes and the efficiency of city services.
The IDB’s Regional Public Good project “Big Data for Sustainable Urban Development”[2] includes five cities in Latin America (São Paulo, Montevideo, Quito, Miraflores and Xalapa) and aims at exchanging experiences and developing new projects based on data for the design of public policies. The objective of this project, presented at the SmartCity Expo Congress in Barcelona in 2018, is to break the natural barrier between policy makers and citizens, and to bring citizens’ demands to the top of the agenda through the analysis of anonymous data and information. For this purpose, the project foresees that the data generated from emergency services centers (911), municipal/governmental applications, and even from the private sector, such as Waze, will be used to support and ensure that political decisions are always supported by up-to-date scientific data and have social foundations that guarantee sustainable development.
To do this, we developed a model that prioritizes road repair needs for the city administrators using citizen-reported data about potholes on roads. And as with many distributions of large amounts of data from larger cities, an unequal Pareto distribution is observed. In this case, the Pareto Principle indicates that about 80% of the reports are concentrated in just 20% of the report locations. The graph below (by João Carabetta) shows us that if a city administrator wants to meet the population’s demand, they should start with the most demanded places. Therefore, if they invest only 20% of total resources needed to correct all asphalt problems reported in the right places, they will serve almost 80% of citizens.
The main challenge in using the data produced by society, such as Twitter, Waze, Facebook, etc., is that it is not as simple as it may seem. First, in order to use the data and compare it to other data sources, we need to validate how representative it is. As studies have shown, citizens do not always have equal access to the data produced. This can be seen in the example below that compares income distributions for the population as a whole and for the population through the lens of a social network [3].
A second challenge is that big data is not produced in a usable format ready for research purposes or for serving the purposes of a city administrator. Usually, working with big data requires extracting valuable information through the use of complex algorithms (as shown below) and an intense effort of structuring, cleaning, and treatment so it can be usable.
Figure 01: raw data of potholes in São Paulo
Figure 02: Aggregated data, but not yet superimposed on the roads
Figure 03: data already added and calculated on the roads
Figure 04: Selection of the top 10% of the road segments with the most reported potholes
Currently, we are addressing the third challenge to be able to locally validate the data generated by citizens in cities. The expected result is that we will be able to develop models that apply to all cities in Latin America and the Caribbean, allowing city administrators to use data to make decision that are based on information gathered from their own citizens.
This is just a glimpse of what is intended to be developed within the framework of the “Big Data for Sustainable Urban Development” project. In the future, we hope to use additional metrics to evaluate the quality of services provided, such as response time for fixing problems, reoccurrence of specific problems, and even the number of people affected by it. With the identification of new sources of data and new applications with implications for public decision-making, we hope to help save city resources, reduce negative impacts on the environment, and contribute to the creation of new practices of public administration that better serve society and lead to public decisions based on the voice of the citizens.
IDB Editors: This article also benefited significantly from the technical inputs and support provided by Mauricio Bouskela, Marcelo Facchina, Hallel Elnir and Sarah Benton.
[1]https://publicadministration.un.org/egovkb/en-us/reports/un-e-government-survey-2016
[2] https://www.iadb.org/en/sector/trade/regional-public-goods/home
[3] Netto, v., Pinheiro, M., Meirelles, J., and Leite, H. (2015). Digital Footprints in the cityscape: finding networks of segregation through Big Data. At https://www.researchgate.net/publication/272408306_Digital_footprints_in_the_cityscape_Finding_networks_of_segregation_through_Big_Data
Leave a Reply