Cities push predictive analytics to combat social ills
Predictive analytics has become a buzzword among smart city officials who are eager to use sensor technology and data to help solve entrenched social problems as varied as crime, homelessness, and illiteracy.
Ambitiously, a panel discussion at a recent Smart City Connect conference in Kansas City, Missouri, was called, “Using data science to fight blight, crime, and potholes.” Yeah, sure, I thought, as I took my seat.
The cynics in the room didn’t speak much. But one person who asked not to be named commented later, “It’s naïve to think that technology could be the panacea for social ills in cities.” Ideas like tracking empty trash bins with sensors to see where homeless people scrounge for food and seek shelter sounds very nice. But those ideas won’t be as effective as a broad, community-based approach that’s based on civic goodwill, the attendee said.
I’ve been of two minds about the value of smart city tech since I began reporting and shooting videos on pilot projects four years ago in Singapore, Montreal, New York, Barcelona, and elsewhere. I’ve seen firsthand that cities can achieve sizable cost savings by incorporating technology. For example, Los Angeles installed LED streetlights that dim automatically on bright, moonlit nights, saving the city $9 million a year. What's more, streetlight poles can have Internet of Things (IoT) sensors and cameras attached in order to monitor crowds, pollution, noise, traffic, and parking spaces. In Singapore, sensors on storm water drains activate tweets to warn residents to move their cars in the event of a flash flood.
Still, isn’t it a bit of a stretch to say a city can use tech to lower a city’s murder rate or improve students' reading scores? That’s putting a lot on the back of technology and data analysis, I thought.
Are you ready for the IoT? Despite massive investment, much of the impact of this transformation is yet to be felt. Learn how to approach IT infrastructure readiness.
Bob Bennett, Kansas City's chief innovation officer, has a different view. “I don’t think we’re putting enough stress on technology,” says Bennett. “We have to figure out ways to use technology to solve big societal problems. We’ve got to make the case to companies that it’s in their interest as well. The profit margins for tech vendors are smaller when working with the public sector, but I need data aggregation for things like what’s happening with education. I’ve got to move faster, because the rest of the world is.”
Bennett talks quickly and passionately. He became a city employee two years ago after a U.S. Army career that took him to Iraq working under top commanders. He has a lot to say, and there’s a long way to go.
Mayor Sly James recently used his state of the city address to bemoan the city’s crime rate, which hit 149 murders in 2017. The mayor also praised an improvement in third-grader reading scores, though he lamented they are still below standard.
Bennett takes the mayor’s concerns to heart. “I want to help,” he says.
“Our murder rate is a catastrophe,” says Bennett. “We’ve got to get to work on it. If I could figure out 10 different factors that correlate with high crime, it would help. I don’t have a hypothesis or a big data recipe. But let me see the correlation between how many kids are on the school lunch programs and crime. Kids who are hungry are probably more likely to commit crimes.”
Big data and predictive analytics are already in use in private industry, often to find consumer preferences for future products such as the next in-demand smartphone features. Bennett wants to make sure the public sector keeps up.
Putting data to work on gunshots
Using existing data about shootings, Kansas City has recently correlated gun violence within the 12-square-block Westport nightlife district within certain weekend hours. That insight helped city officials document and then approve a Community Improvement District to allow restaurant and bar owners to hire private security and off-duty cops to patrol the sidewalks at those most dangerous times, calling on-duty police when needed. “We’re hoping to see an impact in the next three to four months,” Bennett says.
Elsewhere, the city is using pole-mounted video cameras to monitor crowds that leave the Sprint Center arena. People meander through a downtown restaurant area called the Power and Light District, where the city can dispatch added patrols when needed.
“We don’t see much violent crime near Power and Light, but there are drug- and alcohol-related problems,” says Chris Crosby, CEO of Xaqt, the city’s data analytics partner. “We run the video analytics right on the camera box. As we understand the pedestrian movement and flow, it helps us better forecast the crime. With more pedestrians on the street, crime is usually down.”
Dozens of U.S. cities, including New York and Kansas City, have relied on sophisticated ShotSpotter’s gunfire-specific acoustic sensing technology to respond quickly to the distinct sound of gunfire. Such technology won’t necessarily prevent gunfire, but it can lead to quicker arrests and victim response, and possibly shorten the severity of an attack. Kansas City recently issued a request for proposal to add more ShotSpotter technology, although it won’t say for which neighborhoods.
“How likely would ShotSpotter make an impact on our murder rate? I don’t know,” Bennett says. The impact of ShotSpotter will surely become a data point.
Can data analysis improve reading scores?
Improving reading scores for public school third graders sounds like a challenge that could be helped with faster broadband to every home. In San Francisco, an ambitious plan is underway to create a $2 billion public fiber utility to reach every residence, attacking a 12 percent gap in Internet access.
Before Kansas City became the first city in the nation to welcome Google Fiber in 2012, it was estimated that 17 percent of its population didn’t have fast Internet access and more so in poor neighborhoods. There have been obvious improvements in Internet connections over the past six years, but the city of 580,000 residents can’t currently put an actual number on how big its overall digital divide is, even though Xaqt has aggregated plenty of data that can be viewed on a public dashboard. (A separate Xaqt dashboard shows real-time traffic and downtown parking along the city's 2.1-mile streetcar corridor.)
The Smart KCMO digital inclusion dashboard is amazing. The Kansas City region has more than 2 million residents, on both the Missouri and Kansas sides of the river. You can zoom in on an area map to view any census tract and see what level of Internet speed is available, going in five increments from zero to 500-plus Mbps. But while the dashboard gives a good view of pockets of poor Internet speeds and shows how fast speeds are near somebody’s home, there’s not an overall number available for the region's Internet adoption rate. Also, carriers like Google Fiber don’t release subscriber numbers.
It’s not particularly a problem, except when you need to compare Kansas City overall to, say, San Francisco, which says it has 100,000 residents—12 percent of the total—without Internet.
Bennett doesn’t completely buy the notion that fast Internet and computerized learning will greatly improve reading scores, especially for the very young. Kansas City already provides older public school students with computers, and there are computer labs in most schools and public housing complexes. Hometown wireless carrier Sprint has a free Wi-Fi initiative to help underserved communities.
However, the biggest variable in improving young student reading is parental involvement, Bennett and others believe. “Universal broadband is nice, but there’s no silver bullet,” he says. “Governing is really, really hard, and there’s no single software app that fixes all this. It takes a lot of blood and sweat and tears.”
Rick Usher, assistant city manager, says an updated digital inclusion survey in the city is needed, but many factors have to be considered. “We are also recognizing that wireless carriers are serving the [digital inclusion] need as mobile apps become more useful for education, banking, and employment,” Usher says. “It’s hard to just say that Internet connections to a computer at home is the answer anymore.”
Road maintenance reductions by up to 40 percent
One tangible city service that Kansas City has begun to tackle with data analysis is street maintenance. By using data on street conditions gathered from observers, repair records, and a comparative test along lanes on a portion of Ward Parkway, the city has determined when cracks in pavement should be repaired with a sealant instead of delaying until potholes emerge to make repairs—a far more expensive process. The city expects a 20 to 40 percent savings in its road repair budget, covering 6,200 miles of roads with this new preventative maintenance process, Bennett says.
“The savings are not going to be immediate,” Bennett says. Road maintenance is just one area where predictive analytics will help achieve results over time.
But road improvements aren’t as complex as lowering the murder rate. “We have to have a long-range view on these matters. It’s a new way of thinking about local government, a fundamental restructuring of the way we act,” Bennett says.
One consultant’s view of data analysis for city ills
Gartner analyst Bettina Tratz-Ryan, a consultant on smart city initiatives globally, sometimes hears critics question the efficacy of using involved data analysis to solve protracted social ills. Data analysis can’t replace the need for civic resolve.
“Gartner believes that technology should be enabling services that support social and community objectives and approaches, including the ability to solve demographic, aspirational, or contextualized questions,” says Tratz-Ryan. “With the ability to analyze and cross-reference data across urban ecosystems, government officials and startup companies have direct insights into root causes.
“However, not every insight can be solved through technology, nor can it be identified,” she adds. “Not all citizens have the same level of digital activity or maturity. There is always the risk that technology today may only give you the view of the most digitally connected or evident issues. Only with secondary or linked analysis will we find that there are some very analog social problems that we can only solve through analog remedies…like communal empathy.”
Which makes me wonder: Is there a really good app for building communal empathy?
City predictive analytics: Lessons for leaders
- Data science needs to be the bedrock of determining strategy in government as well as industry.
- Employ visionaries and innovators who want to push the limits of analytics.
- Prepare for linking the results from data analysis to comprehensive, futuristic solutions.
This article/content was written by the individual writer identified and does not necessarily reflect the view of Hewlett Packard Enterprise Company.