Urban researchers from MIT and Harvard are now using a computer vision system developed by MIT Media Lab to identify factors that can predict urban change and quantify the physical improvements as well as the deterioration of neighborhoods in five U.S. cities. The tool is designed to predict changes that will happen in coming years with the aim of helping municipalities take measures accordingly. “The combination of high-quality measurement, analysis, and thoughtful attention to what is missing is the future of measurement,” comments Julia Lane, a professor at New York University’s Center for Urban Science and Progress.

neighborhoods

Courtesy of MIT and Harvard researchers.

The system is composed of images of neighborhoods and compares 1.6 million pairs of photos taken periodically for seven years. The researchers use the results of those comparisons to test and challenge several hypotheses by social scientists concerning the causes of urban revitalization. They find that density of highly educated residents is strongly correlated with the physical improvements of a neighborhood. However, it’s not about the level of wealth, rather the skills coupled with the level of education of these neighborhoods that determines their revitalization.

They also prove a theory wrong: that when a neighborhood has buildings in an advanced state of decay, residents begin renovating. According to the researchers, that relationship isn’t necessarily true for the neighborhoods that they studied. Another theory was accurate in practice though: neighborhoods that are doing well are more likely to do better, while those not doing so well won’t be doing any better.

neighborhoods

MIT.

However, the images may not tell the whole story. In 2015, the City Observatory conducted a study that found that, from 1980 to 2010, the proportion of lower-income households living in a low-income neighborhood increased from 23 to 25%, whereas the proportion of upper-income households living in an upper-income neighborhood doubled, going from from 9 to 18%. It would be interesting to see how neighborhoods have changed, since – according to the researchers – it’s not the income of the neighborhood that changes it to the better or worse.