Smart Cities Research - 2017

Using New Technologies to Survey Memphis Residents: Identifying and Responding to local concerns

Dr. Michael W. Sances

Another way to find out how citizens think and feel is to ask them. That’s the idea behind the traditional public opinion poll. However, according to Dr. Michael Sances, also of the Smart Cities Cluster, traditional phone polls aren’t just old-fashioned; they are cost prohibitive for city governments. “Yet knowing how citizens feel about key issues can make all the difference in delivering services and increasing quality of life for residents,” he says. Sances’ tool of choice? Facebook. More specifically, the Facebook Ads platform, which enables a city to create targeted advertising for an online poll. According to Sances, Facebook estimates that 490,000 Memphians are reachable via Facebook, or about 75 percent of the 2015 population. “The potential for responses is high compared to conventional phone polls,” he says. “The viewer sees the ad, clicks on it, and is taken to a questionnaire. It’s targeted. It’s affordable. And it holds a lot of promise for success.” The study will compare the results and the costs of the online polling with those of previous phone polls conducted by the city of Memphis.

Poster | Presentation | Report

Integrated Healthy Home Assessment and Intervention for Children in Memphis

Dr. Pratik Banerjee

Asthma and lead paint are common threats to pediatric health in Memphis, and they often go hand-in-hand. But seldom have they been addressed together in an integrated intervention program. Dr. Pratik Banerjee of the Institute of Intelligent Systems and the Smart Cites Cluster is poised to do just that, equipped with new tools for the job. The National Institute of Occupational Safety and Health has developed a special bioaerosol collection and assessment method and made it available to the University of Memphis School of Public Health. “We have exclusive access,” Banerjee, says. “So I’m converging this method with HUDapproved protocols for lead and indoor allergens sampling to develop a robust, integrated health homes assessment regimen that properly addresses asthma and lead paint issues together.” The new assessment regimen will offer a more complete picture of a home’s health issues so that a more comprehensive solution can be provided efficiently and effectively.

Poster | Presentation | Report

Linking Home Energy Insecurity to the Built Environment and Population Health in Memphis

Dr. Chunrong Jia

Just as asthma and lead poisoning haven’t been adequately addressed, neither has the impact of energy insecurity on these and other health issues. According to Dr. Churong Jia of the Smart Cities Cluster, the inability to adequately meet household energy needs is often accompanied by food insecurity, poor health, high rates of hospitalization and developmental risks. “These low-income householders are often faced with a ‘heat or eat’ dilemma,” Jia says. Jia is working to assess the built environment, energy insecurity, and health outcomes together to position energy insecurity as a major public health concern. Eventually this work could lead to scalable policy changes that could improve energy efficiency, reduce exposure to adverse conditions, and improve health and wellbeing in socioeconomically disadvantaged households.

Poster | Presentation | Report

Modeling Adoption of Technological Innovations and Infrastructure Impacts in a Smart City

Dr. Sabya Mishra

As technology advances, transportation will change forever. But the transition from conventional automobile travel to a world of autonomous vehicles and smart travel infrastructure won’t be instantaneous. Or easy. That’s the impetus behind the research of Dr. Sabya Mishra of Smart Cities. Up to now, existing methods of analyzing travel behavior haven’t taken into account emerging technologies. It is Mishra’s goal to incorporate smart and connected city infrastructure in the analysis of travel behavior to see how it changes when users of new technology share the road with those still using traditional technologies. “We hope to be able to predict the probability of citizens adopting innovative transportation technologies and, as new technology is employed, see how behavior changes in a wide range of circumstances, including travel for work, shopping and leisure,” Mishra says.

Poster | Presentation | Report

A New Chapter for Recycling: Using Geographic Information Systems (GIS) and mixed-methods approach to improve the market development and transportation of recycled materials

Dr. Esra Ozdenerol

Dr. Esra Ozdenerol of the Smart Cities Cluster is interested in using the technology to bolster recycling in Memphis. “Improving the market development of recyclables in Memphis is a complex and challenging task,” she says. “Making it less costly and more rewarding is imperative.” Her research includes interviews with restaurant owners, apartment complex managers, private recyclables collectors, public works employees and others at the state and federal level to collect relevant recycling and transportation data. It also includes teaming up with a private collector and designing a GIS-based plan that can determine “least distance” and “least cost” routes for collecting recyclables and getting recycled material from processor to manufacturer. “By determining optimal transportation routes,” says Ozdenerol, “we can begin to develop an ideal recycling infrastructure for the city.”

Poster | Presentation | Report

A Planning Support System for Comprehensive Planning and Zoning: A Geospatial Simulation Model of Land Use, Land Cover Change for the Memphis Metropolitan Region

Dr. Reza Banai, Dr. Youngsang Kwon

“The menacing consequences of humanmade urban sprawl are comparable to those of the naturally occurring disasters like floods, hurricanes, and earthquakes,” says Dr. Reza Banai. That is why he and Dr. Youngsang Kwon of Smart Cities are using GIS to map how the natural and urban landscape changes over time. Of particular concern is how the metropolitan region is particularly vulnerable in built-up areas near rivers, creeks and wetlands. “These areas become impervious and pose increased risk of local flooding during intensive- rain events in neighborhoods, damaging both the ecosystem and personal property,” says Kwon. By mapping what’s called Land Use/Land Cover Change or LUCC, particularly as it transforms from undeveloped land to urban development, and by looking at other indicators such as per capita consumption of land, Banai and Kwon’s study will help inform smart long-term planning and zoning for future smart growth and environmental sustainability.

Poster | Presentation | Report

Predicting Localized, Fine-Grained Crime Types using Twitter

Dr. Deepak Venugopal

The Windsors aren’t the only researchers who are trying to tap into the predictive power of social media. Dr. Deepak Venugopal of the Smart Cities Cluster says Twitter has more than 100 million users generating more than 300 million tweets each day. Subject matter covers a wide range of topics, and he believes these messages can be mined for information that can help our communities fight crime. According to Venugopal, Twitter data has already been used successfully to predict election outcomes, predict disease outbreaks (such as the spread of flu in a specific community), and more. “Why not use it to predict crime?” he asks. Venugopal’s goal is to build an information extraction system that can assess a wide variety of tweets, from crime reports by news agencies to tweets expressing hate or other emotions related to destructive and violent behavior, and map them to reveal locations that have a higher probability for crimes of various types. This information will be combined with data from local law enforcement agencies to build an adaptive fine-grained prediction model that will help the city of Memphis allocate expensive law enforcement resources to areas that need it most.

Poster | Presentation

Detection of Unusual Objects, Actions and Events in Streaming Video Surveillance Data

Bonny Banerjee

We present an online sparse coding based approach for abnormal event detection in videos. The approach is based on the intuition that abnormal events occur rarely. It extracts local spatiotemporal volumes and uses an online sparse dictionary learning algorithm to learn a set of atoms. While learning, the rarity of atoms is approximated online using Incremental Coding Length to measure the entropy gain of each atom. To speed up the process, a batch version of the well-known Orthogonal Matching Pursuit algorithm is used. The proposed approach operates in an unsupervised and online manner, hence applicable to real world streaming video. Experiments on three benchmark datasets (UCSD, UMN and Subway) and evaluations in comparison with a number of mainstream algorithms show that the proposed approach is comparable to the state-of-the-art.

Poster | Presentation

Uncovering Latent Community Issues Through Social Media

Dr. Leah C. Windsor, Dr. Alistair Windsor

It’s an unfortunate reality. Cities have a difficult time knowing what their citizens want or need until they complain. Many crimes go unaddressed because they are never reported. Service issues often take weeks or longer to resolve, simply because no one filed a formal complaint. We do have 311. It allows Memphians to call in and request services, such as curbside debris pickup. But what if the city could know sooner about existing blight, litter and crime? Drs. Leah and Alistair Windsor of the Institute of Intelligent Systems and the Smart Cities innovation research cluster are exploring how social media can play a role. “Because there is a lack of formal reporting, citizens feel like their needs are not being met,” explains Leah Windsor. “So they turn to social media like Twitter to alert friends and neighbors about local problems.” The Windsors are probing existing social media outlets with the hope of eventually building an automated tool that can collect and analyze information from these sites to detect latent problems, reveal patterns in problematic areas, and track emerging patterns in new areas. With that kind of information, says Alistair Windsor, “Community resources can be better allocated, increasing government efficiency and improving the citizens’ relationship to their neighborhoods and their city officials.”