Release: Smart City Evolution: Roadmap to Vision Zero

(This post originally appeared on Microsoft CityNext, by Jeff Friedman, Director of Modern Government, Microsoft Corporation on August 16, 2017)

Titled “The Smart City Evolution: A Roadmap for Vision Zero,” the paper is aimed at revealing common challenges, identifying best practices, and exposing opportunities for cities to use data science to inform traffic management strategy and tactics. With support from Microsoft, ODN conducted comprehensive research, interviewing administrators from 25 municipalities across 23 U.S. states and performing detailed data surveys in 40 municipalities with populations ranging from 46,000 to 8.5 million. The goal, according to ODN founder Carey Anne Nadeau: “Trying to better understand what cities can do to prevent crashes and save lives by leveraging local data to inform their decisions, optimize the impact of their budgets and prepare for autonomous vehicles.”

Through this extensive research, Nadeau said they recognized an important insight: not all smart cities are alike in their quest to Vision Zero. “They’re on this journey … of becoming more data-minded and technologically sophisticated,” she explained. “This paper acknowledges that there are commonalities across (one of four) different stages we identified,” including:

  • Initiative: Developing an action plan with community engagement.
  • Descriptive: Analyzing and visualizing the magnitude of the problem.
  • Active: Responding with activities in known problem areas.
  • Adaptive: Planning and learning how and where to optimize for impact.

U.S. cities can use the white paper to self-identify their stage and choose appropriate strategies to mitigate traffic deaths and injuries—from organizing executive and community buy-in to preparing data for statistical models and autonomous vehicles. With 2016 proving to be the deadliest year on U.S. roads since 2007, the paper’s timing is critical. “Unfortunately, many cities across the United States are seeing increases in the number of traffic crashes and fatalities in their communities,” Nadeau said. “Crashes cost $62 billion a year, not considering the loss of human life. We think of this as a very important, deepening and really dire issue for cities.”

Using Azure to prevent traffic injuries and fatalities

Microsoft is making our trusted, secure Azure cloud platform available so that ODN can curate a database of U.S. city information that is already collected—from police, fire and emergency management reports, and transportation, permit and inspection records—and use data science to predict crashes before they happen, preventing injuries and saving lives. 

The idea is to use machine learning to reveal patterns and techniques that produce the best results. “A number of data sets are floating around (in cities),” Nadeau said. “What Microsoft Azure allows Open Data Nation to do is bring together data from disparate sources … and determine through both qualitative surveys and machine learning what variables are important predictors of traffic crashes.”

In addition to educating cities about the value of leveraging their data to prevent traffic deaths and serious injuries, Nadeau said the white paper also provides insights for autonomous vehicle manufacturers: “Autonomous vehicles will be introduced into cities in the next five years. (The white paper provides) an education of the autonomous vehicle industry to realize that city data is the way to make their safety features more robust.”

Eliminating all traffic crashes

In closing, Nadeau reinforces a key message to city leaders: “We don’t have to wait for a traffic crash to happen to act; it is possible to eliminate all traffic crashes. We make that seemingly improbable outcome possible by taking advantage of machine learning to predict what is going to happen before anyone is injured or killed in a traffic crash.”

Microsoft CityNext is proud to join ODN in using the cloud and data science to plan and prioritize traffic-safety resources more effectively and efficiently. To learn more, please visit: