By Data-Smart City Solutions • June 1, 2018

Each week we will bring you a summary of what happened this week on our site, on Twitter, and in the wider world of civic data. Suggest stories on Twitter with #ThisWeekInData.

IEEE Spectrum reported on a dispatching algorithm created by researchers at MIT that could cut New York City’s taxi fleet by 30 percent. Published in Nature, the research sought to minimize the distance between a taxi’s destination and the origin of its next potential trip, moving more passengers per vehicle over a given period of time. Of course, while good for congestion, the implementation of this model could present another hit to taxi drivers, many of whom have already been replaced by ridesharing apps.   


MIT Technology Review profiled a new tool under development by Microsoft that can automatically identify bias in a range of different AI algorithms. Concern has recently risen over the presence of bias in algorithms used by governments and private companies alike to make decisions ranging from sentencing to hiring. This tool presents an automated way of detecting unfairness in models before they’re deployed, and follows a few weeks after Facebook announced Fairness Flow, its own tool for detecting bias.  


Bloomberg Philanthropies announced an additional $42 million investment in What Works Cities, a program to enhance cities’ use of data and evidence to improve resident outcomes and address the most pressing local issues. What Works Cities has already partnered with 100 mid-size U.S. cities to embed data-driven solutions into municipal practices, and will continue with its WWC Certification Program—which establishes standards and evaluates cities on their data-driven success—as well as grow its network of partners.


CityLab outlined a project deployed by the City of Austin that uses blockchain to create a portable, digital identity for homeless people. Individuals often require birth certificates, social security cards, and health insurance records to be connected with homelessness services, yet many homeless people lack this documentation. By hosting information on the blockchain, Austin can ensure personal records of Austin’s homeless will be encrypted, digitized, and impossible to lose. While city officials are the ones facilitating the recording of this information, individuals themselves will have access to their own records, and control over who can see it.


Route Fifty analyzed the City of New Orleans’ initiative to leverage administrative data in order to improve quickly rising ambulance response times. The city’s EMS staff partnered with New Orleans’ Office of Performance and Accountability and students from Louisiana State University’s Masters of Science and Analytics Program to answer two questions: how EMS ambulances were picked to address a 911 call and where they were stationed after responding to a call to wait for the next one. Analysts plotted every EMS call citywide on a data visualization and identified over 100 potential waiting locations that would create eight-minute response times for high- and low-traffic conditions, producing printable maps indicating optimal locations for EMS ambulances to wait for their next call.


Here on Data-Smart, Jess Weaver highlighted strategies for organizing and operationalizing the flood of data that comes during emergency situations. Recommendations include standardizing data so it can be easily shared across departments, pursuing predictive analytics projects to inform preemptive action, and implementing analytics techniques to spot high priority data among the influx of information.


The Stanford Social Innovation Review outlined ten reasons not to measure impact and discussed what to do instead. The authors argue that while high-quality impact evaluations have immense value, a great deal of money and time has been wasted on poorly designed, poorly implemented, and poorly conceived impact evaluations that lack quality data or fail to engage stakeholders.  The push for more and more impact measurement can produce poor studies and wasted money, as well as take resources from collecting data that can actually help improve performance. The article argues that organizations need to find “right-fit” evidence strategies that pay careful attention to context.


GovTech discussed efforts by the State of Texas to improve digital experiences for residents. The state contracted Deloitte earlier this year to develop its online platform, and will soon launch a new personal digital assistant. Even more impressive, the arrival this fall of the My Government, My Way initiative will enable a citizen portal through that will let residents set up a personalized online service experience.


Living Cities examined efforts by a diverse group of community members in San Antonio to embed data into day-to-day operations. While agencies in San Antonio have long relied on data and evidence, a push from the Mayor’s Office has helped create a citywide culture of data. By convening a brainstorming session with 50 people across the city government, the Mayor’s Office was able to understand the various perspectives on data-driven work and then follow up on these ideas via a steering committee supported by public employees across the city enterprise. 


GovTech profiled the State of Montana’s suite of chatbots that help residents navigate the online services provided by the DMV. Created almost entirely by a single employee in the Department of Justice, these chatbots have not only enhanced user experience, but also reduced the volume of problems that have to be sorted out by staff at the DMV. The employee, Levi Worts, has no background in coding, but spent months reading through documentation and best practices while also teaming with the user-experience company Tars, which offers a chatbot builder for inexperienced users.


On Governing, Stephen Goldsmith analyzed ways that cities can use data and technology to prepare residents for a quickly changing workforce. With troves of data from social media companies like LinkedIn, cities now have the opportunity to analyze job openings in near real-time. Around the country, cities are leveraging market analytics to align labor supply with demand, develop better training programs, and work with institutions to fill education gaps.