Guide to Building Data-Driven Organizations in the Public Sector

Remote Sensors

Team 2: Matthew Simon and Carlos Lopez

Topic Overview

Technology is changing the way we do business, the way we travel, and the way we live. In this week’s readings the authors include interesting perspectives regarding how a data driven city could operate to create a more productive, healthier, and engaging society. For example, both authors touched on transportation and the available opportunities with GPS data in our phones. Basically, when one sits in a traffic jam, Google can capture the minute by minute data on our phones and when combined with the data from other drivers it is able to accurately and anonymously demonstrate traffic conditions and delays (Eagle & Greene, 2014).

Google Maps is a great tool, one that is anonymous, and can be used daily for commutes. Alex Pentland in “Social Physics” offers the “smart city” and takes it to the next level where your commuting patterns along with the rest of the population in your city are part of a model that can provide you with the optimized time and route for your trip so that you have the most efficient travel experience on your way to work. Similarly, commercial vehicles can identify the optimized travel times reducing conflicts with passenger vehicles and improving their efficiency. Further, Pentland describes a “smart city” where people getting the flu and their recent whereabouts could be mapped and when that is overlaid with others that also have the flu the location of where the flu started could be identified and contained before it spreads further.

Smart cities may be the future or perhaps elements of this are already in place. We produce data every day such as how long we are on the road, what we ate for lunch, and how many steps per day we take. The authors basically propose a smart city that takes the individual life logging efforts per se that could be documented on our phones and when combined with the life logs of the rest of the population it can clearly demonstrate how the city operates, identify energy and transportation peaks and valleys, improve emergency services and security, and develop a city that works for the people living in it. The overarching concern and what seems to be the theme for this course is privacy. The authors explain the need for privacy and customer buy-in for long term success; however, if there are data breaches where data is comprised it would deter and setback the smart city effort back to more conventional ways.

Chapter Summaries

Eagle, N., & Greene, K. (2014). Reality mining: Using big data to engineer a better world. MIT Press. CH5 urban analytics: traffic data, crime stats, and closed-circuit cameras

This chapter begins with an intriguing problem statement: The cost of congestion is exceeding $100 billion per year and wasting 34 hours per commuter per year according to the 2011 Urban Mobility Report. The conventional ways to acquire a morning or evening traffic reports included the local radio station but now there are applications such as Google Maps, cameras on the freeways that can be viewed online, and websites that can identify any accident or construction delays.

Eagle and Green provided examples of companies seeking to provide a better traffic report service such as Intirx – which aggregates a variety of traffic data to make sure navigational devises in people’s cars are up to date. Intrix has partnered with Audi, Nissan, and Ford to provide traffic data for their cars’ built in navigation systems as well as the University of Maryland and the Interstate 95 (I-95) Corridor Coalition. The 20,000 miles of highway traversing 10 states along I-95 from Florida to Maine has helped Departments of Transportation determine where and how to better allocate transportation resources.

This is a similar issue facing the Arizona Department of Transportation where the revenue forecasts allows for the preservation and maintenance of the existing highway system with little or no funding to expand outside of the metro regions of Phoenix and Tucson.

Eagle and Greene also discuss data for predicting crime. At the core of predicting crime is a solid database of the crimes that have occurred previously. Incident and arrest reports including the time, date, location, crime code, and persons involved provide data that help the department keep a history and trends in their region.

The next generation of data for predicting crime includes mapping and providing all of this information in a visual format. More recently, specialized algorithms and real time crime data is frequently updating crime maps that can position police in locations before crimes occur. One example is in Memphis, Tennessee where police have run a program since 2005 called “Blue crush” (Criminal Reduction Utilizing Statistical History) that examines current activity levels and shifts in crime levels due to previous changes in police coverage (Eagle & Greene, 2014).

Lastly, Eagle and Greene touch on video data to catch and possibly deter criminal activity. The Department of Homeland Security (DHS) has funded many police department’s cameras in an effort to combat terrorist threats; in 2009 DHS spent $15 million in seven cities. In 2010, more than $830 million went to 64 metropolitan areas; in 2011 31 cities used $662 million of funding under their initiative. Research is mixed about the effectiveness of cameras. Studies in Los Angeles, London, Chicago, and Baltimore found some areas to have little or no effect in deterring criminals and in other areas such as Chicago crime dropped 12%.

Pentland, A. (2015). Social Physics: How social networks can make us smarter. Penguin. CH8 sensing citiesThis chapter focuses on two types of data thac can define the rhytm of a city: traffic metrics and crime statistics

This chapter sets the current conditions of how cities currently operate and offers a concept of what they could be if cities adopted data-driven and “smart city” initiatives.

Pentland sets the stage by going back to the 1800’s when the industrial revolution spurred rapid urban growth and created huge social and environmental problems. The remedy then was to build centralized networks that delivered clean water and safe food, enabled commerce, removed waste, provided energy, facilitated transportation, and access to health care, police, and education.
However, the author points out that these solutions are outdated and becoming “increasingly obsolete” as cities struggle with transportation, health care, and education issues and more. Pentland offers a different framework of rather than having static systems that are separated by function – water, food, transport, to consider them as dynamic and holistic.
What would be the data source for these dynamic and holistic efforts to take shape? You guessed it via the mobile phone. Pentland explains that wireless devices and networks could become the eyes and ears of an all controlling “smart city.”

But how does this happen?

First, this requires social physics, specifically the visualization tools that will allow citizens to use these new data streams to manage the city. The visualization will inform citizens about where people live, where industrial/commercial areas are located, and the demographics of its residents. The second step the author outlines is a new deal on data, an architecture and legal policy that guarantees privacy, stability and efficient government.

Pentland provides and emphasis on Behavior demographics. Pentland explains,

“For most people, the primary pattern is the workday, that is, going to work and coming home, usually along the same path day after day. The second most pronounced pattern is the weekend and days off, often with the characteristic behavior of sleeping in and spending that night out in a location besides home or work. Perhaps surprisingly, the places we go and things we do during our free time are almost as regular as our work patterns. The third patters is a wild card – days spent exploring, usually a shopping trip or an outing; together these three patterns typically account for 90 percent of our behavior. In summary, by combining these habits in time with the behavior demographics it can allow us to better plan city transportation, services, and growth; these data driven forecasts allow us to prepare for peaks in demand and manage them better. The ability to know where and when the people who are at risks of diabetes eat, or where the people who have trouble handling money shop also has great potential to improving public health and education.”

Health issues could also be diagnosed based on individual behavior. Pentland describes that people feeling sick tend to behave differently; for example, those with a sore throat and cough symptoms were found to have their normal pattern of socialization disrupted, and they began to interact with more and different people. Those with a common cold, their overall number of interactions and nighttime interactions increased. People with fever limited their movement and people feeling stressed, sad, depressed became socially isolated on symptomatic days.

Behaviors can be signs of an emerging illness and the author points out an app such as Ginger.io that could identify that change in behavior and figure out if an illness is coming. To take it a step further, the author explains that by crowdsourcing this behavior across a population and then combining that info with data about where and when people went the infection risk area can be figured out. If this was known then action could be taken to avoid spreading the disease further.

Key Take-Aways (for Yellowdig)

Discussion Questions

1) Traffic, crime, and health technology are rapidly improving to better predict and identify incidents. What opportunities or concerns do you find with technology and the data driven approach?

2) What are examples of emerging technology in your industry? What are the strengths and/or weaknesses?

3) Would you be interested in living in a “smart city” as Pentland describes? What would you change or improve?

References