Pitt Researchers Develop New Science of National Preparedness
It could be said that nothing shook the nation to its core—and catalyzed a transformation in the way the nation thinks about the complexities of national security and disaster preparedness—like the tragedies of Sept. 11, 2001, and Hurricane Katrina in 2005.
Certainly, preparedness discussions already were occurring at places like the University of Pittsburgh long before those infamous events. But amidst those disasters, the nation’s lawmakers, law enforcers, policy analysts, constitutional rights advocates, public health officials, first responders, the military, and academia suddenly found themselves front and center in national and regional debates. With a new sense of urgency, they grappled with the future what-ifs of disasters—how to better prepare for them, anticipate them, manage them, and mitigate their consequences.
An intellectual “Code Orange” thus ensued over the coming years as thoughtful leaders around the country theorized, planned, and simulated possible future disasters—and worked to solve technological problems at a level of intensity not seen before.
Among them, of course, was a diverse group of University of Pittsburgh researchers and policy experts who joined
forces to hold their own academic discussions, conduct intensive research, and, ultimately, develop science-driven, big-picture frameworks to harness the complexities of large-scale disasters. That collective initiative culminated in 2004 with the launch of the University of Pittsburgh Center for National Preparedness.
Since then, the multidisciplinary center has helped drive the national agenda. Its leaders hail from Pitt’s Schools of Medicine, Information Sciences, Engineering, Law, and Nursing, as well as, among others, the Kenneth P. Dietrich School of Arts and Sciences and the Graduate Schools of Public and International Affairs (GSPIA) and Public Health (GSPH). The center’s researchers have developed new innovations in such areas as disaster management, robotic search and rescue, information analysis, and public health monitoring.
But that’s only the beginning.
By taking the issue of national preparedness to a new level, these collaborative experts have created a whole new academic discipline at Pitt.
“National preparedness is about having leaders who expect or anticipate things that previously were unimaginable,” says Carey Balaban, a professor in the School of Medicine’s Departments of Otolaryngology and Neurobiology and a seemingly unlikely codirector of the Center for National Preparedness. “These leaders know the possibilities, try to mitigate them beforehand, respond quickly when they occur, and build to prevent them from recurring. It’s a cycle.”
And now it’s a science, too, according to Balaban and Kenneth Sochats, an information systems engineer who also is a founding codirector of the center.
“We’re making a science of national preparedness,” Balaban declares. “We have taken an evidence-based, systems-of-systems analytic approach to issues of national preparedness. We are bringing the full rigor of the academic endeavors to practical problems that improve outcomes for the good of society.”
A brief history
While Pitt researchers had been studying national preparedness issues substantively prior to Sept. 11, 2001, the fall of the World Trade Center towers quickly prompted the University’s Office of the Provost to evaluate its own potential for creating a national preparedness-focused research program, which included an inventory of any current, related research.
A consulting firm’s conclusion at the time, according to George Klinzing, vice provost for research at Pitt: “You already have one. The Scharpenburg Report [as the consulting firm’s report came to be known] confirmed that we have great strength in the areas of health and information science. The knowledge is here. Now we’re pulling it together.”
And they did. But as Klinzing noted at the time of the launch of the Center for National Preparedness, “This is not just about homeland security. We’re finding that the different aspects of national preparedness add up to a much larger picture.”
Disease outbreak surveillance and the national agenda
As the consultant pointed out, Pitt found strength—and much favorable national media attention—largely in its collaborative mix of public health and information sciences research. In early 2002, for instance, Pitt became a poster child for the national agenda on preparedness and security when then-President George W. Bush and his Department of Homeland Security director, former Pennsylvania governor Tom Ridge, visited Pitt. Their focus, which at the time became an integral talking point within the president’s national agenda: a Pitt research initiative called RODS, or Real-time Outbreak and Disease Surveillance.
RODS is an information system that was designed by Michael M. Wagner of the School of Medicine’s Department of Biomedical Informatics to immediately identify sudden jumps regionally in emergency room visits related to specific complaints. The system, which captures and monitors such activities, was designed to detect disease outbreaks, with the idea that such spikes could serve as a first warning of a possible bioterrorism event. The RODS research team also had worked on the development of a National Retail Data Monitor, which would collect and analyze data on the sale of over-the-counter drugs—also designed to detect disease outbreaks.
At the same time, a Pitt research team was developing a decision-making process for first responders that included recognition, protection, decontamination, triage, and treatment. That same team embarked on developing what it called the Pittsburgh Matrix, which measured survivorship and cost over a timeline beginning with the detection of a biocontaminant as well as the scale of medical resources required to respond.
At the same time, Margaret Potter, a Pitt professor of health policy and management and director of Pitt’s Center for Public Health Practice, was advocating with the Pennsylvania legislature for a statewide public health communications network that would improve the response to local emergencies. Potter also served as principal investigator with the Graduate School of Public Health’s Center for Public Health Preparedness, one of 22 such centers nationwide that were funded by the Centers for Disease Control and Prevention.
Today, that center (www.prepare.pitt.edu) continues to thrive, educating and training public health workers and school personnel in preparedness issues such as emerging infectious diseases, disasters, preparedness law and policy, and crisis leadership. The center also oversees a graduate certificate program in public health preparedness and disaster response.
On the ethics front, the Matthew B. Ridgway Center for International Security Studies within Pitt’s Graduate School of Public and International Affairs began to study the ethical dilemma presented by effectively gathering security intelligence on potential terror threats while still respecting the constitutional privacy rights of U.S. citizens.
Another initiative that brought Pitt to the forefront of national preparedness shortly after the 9/11 tragedy included the following. Regional Biocontainment Laboratory. The National Institute of Allergy and Infectious Diseases awarded Pitt a $17.5 million grant in 2003 to establish this laboratory—one of nine in the country. The lab is housed in the University’s state-of-the-art Biomedical Science Tower 3.
Preparedness today: an evolving discipline
Today’s Center for National Preparedness continues to pave new paths in helping to drive the nation’s preparedness agenda.
“It has morphed in a lot of ways,” Klinzing says of the center. “I think that, after a broader focus earlier in its evolution, we’ve found some solid niches on which to build. I’m happy with them; we have the right talent.”
Sochats, who has spent more than 30 years working in both academia and the telecommunications industry, says the center has matured in its focus since “the early days, when everybody was scurrying around exploring a number of competing theoretical approaches to managing disasters.”
He adds: “It’s still a new field, but now we’re actually developing tools and producing educational programs.”
And the center’s doing so in partnership with, among others, the Potomac Institute, Harris Corp., Lockheed, the military, federal agencies such as NORAD, FEMA, NIST and NORCOM, and numerous other universities, state agencies, and regional preparedness consortia.
“With the new funding picture nationally,” Sochats says, “we need to partner very closely with the commercial sector, where they actually make things.”
What follows are descriptions of some of the center’s research endeavors, which build largely on the same academic strengths that had been identified in 2002.
An unlikely partnership in disaster management
One could fairly describe Balaban and Sochats as the most unlikely of research partnerships. Balaban is a prolific and frenetic thinker and idea man with a background in medical science and a passion for neuroscience. Sochats is an electrical engineer with a penchant for visual information systems, electronic record keeping, and practical, buildable solutions. The pair originally was brought together by Klinzing as part of a brainstorming team to explore the establishment of the center.
As Klinzing is quick to acknowledge, “They’re just two of the most creative people I know. They really work great together.”
Once together, though, this emerging team not only agreed to lead the new center, but they also began to explore—over lots of Starbucks coffee—the need for new decision-making tools in managing disasters that account for the many “actors” responding to or affected by a disaster and the fast-changing dynamics of the disaster scenario. Eventually, their collaboration led to the development of what they call their Dynamic Discrete Disaster Decision Simulation System, or D4S2.
The patent-pending system, which continues to evolve, integrates a geographic information system, discrete event simulation, a custom-built decision-modeling system, and a control interface that resembles an emergency operations center. It allows users to overlay all actors in a given disaster and informs each group continuously as situations change and decisions are made.
“It’s all about situational awareness,” Balaban says, as he compares the system to the human body’s neural system. “It’s all about prediction and dealing with complex interactive networks and how they operate together. I think it’s actually a very powerful platform for other applications … We expect this to become a premier tool in emergency response.”
Balaban and Sochats are working with the University’s Office of Technology Management to commercialize their innovation.
Avoiding the “cascade of failure”
When Louise Comfort, a professor in GSPIA since 1984, looks at a disaster, she sees multiple actors and situations and a sequence of decision points that can, in cases like the Hurricane Katrina disaster, lead to a “cascade of failure.”
“I’m very interested in the decision making,” says Comfort, an organizational design theorist and policy analyst. “I believe that the decisions made in the initial response to a disaster will set the trajectory of the rest of the situation and determine whether it will escalate or not.”
Take Hurricane Katrina, for example, which Comfort has studied extensively over the years. “It’s incredible that all four levels of government failed” in handling many aspects of the disaster response. “One of the critical issues was a lack of understanding of the scientific information available and how this affected their decisions. There was no capacity to imagine what would happen to the aging infrastructure of the city under the impact of the interconnected dynamics of a storm surge from a Category 3 hurricane flowing through the Mississippi River Gulf Outlet into the low-lying districts of New Orleans.”
Her biggest complaint: “The information was there, but it was very complex, and nobody put the pieces together. Nobody saw the whole. That’s exactly what we’re trying to facilitate with computational decision support for human managers.”
Comfort, who also spent time recently in Japan studying the response to the March 2011 earthquake, tsunami, and subsequent nuclear disaster there, has turned her analysis into what she calls an Interactive Intelligent Spacial Information System (IISIS). She describes her innovation as a computer-based decision support tool that tracks and monitors the interactions between critical conditions, actors, and agencies in real time, providing decision support for emergency managers in a rapidly changing, urgent environment.
“I study ways that computers can extend human capacity for decision making in urgent situations, when human decision-making cognitive capacity drops under stress,” Comfort says of her innovation and research. “The one constraint is that human minds are much faster than machines, but human beings also make more mistakes.”
Comfort says she continues to update her computer system, recently adding modules for an engineering dashboard for hospitals and collaborative decision support for airport fire-rescue teams. Meanwhile, she also is studying the concept of regional risk assessment and is working on helping others “to look at the whole set of hazards that could happen in a region” and helping them develop “complexity profiles,” among other tools, she says.
Is she obsessed with studying disaster response? Maybe a little bit. “I am thinking about disasters all the time,” Comfort deadpans. “But given my teaching schedule, I can probably manage [researching] one disaster a year.”
Making opinions matter
While some researchers work on ways to prevent or manage disasters and possible security threats, one Pitt professor and her research team are trying to harness the potential intelligence that could be extracted and interpreted from vast amounts of printed text.
Janyce Wiebe, a professor in the Dietrich School’s Department of Computer Science and codirector of a multiuniversity Intelligent Systems Program initiative, is collaborating with Rebecca Hwa, a professor of computer science at Pitt. They, along with researchers from several other universities, are working to develop what Wiebe describes as machine learning-based statistical models that can rapidly process large volumes of unstructured text in search of opinions, general sentiments, motivations, and tensions.
The computational models are being designed to automatically merge facts and entity relationships across sets of documents and populate large databases with information from many text sources as they relate to such events as terrorist incidents or disease outbreaks. The design challenges, Wiebe says, are complicated, particularly when considering the researchers’ goal of developing retrainable, robust components for those models.
“To make decisions, policy makers have to be aware of what people are writing and saying,” Wiebe says. “What are all of the opinions being expressed, and how do they travel over time and forums? There are vast amounts of texts, and humans can’t possibly read it all, so we want to develop systems that can bring the relevant data to analysts’ attention.”
And what are these systems looking for in those texts? “I’m looking for subjectivity—the linguistic expression of somebody’s opinions, sentiments, emotions, evaluations, beliefs, speculations,” Wiebe says. “But you can’t just look for points like ‘good’ and ‘bad.’”
Multiple-robot search and rescue simulation
Also building on Pitt’s strength in information sciences is Michael Lewis, a professor of intelligent systems programs in the University’s School of Information Sciences. He started working on human-robot interaction research in the area of search and rescue beginning in 2002, supported by a National Science Foundation Information Technology Research grant.
Lewis, in collaboration with Carnegie Mellon University researchers, began to develop robots, design interfaces, and equipment, all aimed at search and rescue with multiple robots in extreme environments. Ultimately, the researchers developed an urban search-and-rescue simulation that eventually was used in a national Virtual Robots RoboCupRescue Competition. The simulator also has been used by many researchers across the country to support and test their own robotic development efforts in search and rescue.
Among the challenges that Lewis and his collaborators tackled with the simulation platform: organization and command and control using multiple platforms, Lewis says.
“When you have four or five robots in one area, you get lots and lots of redundancy, and it gets very confusing,” Lewis says. “They may be good for search and rescue of static targets, but it’s more difficult with dynamic targets.”
More recently, Lewis and researchers from Carnegie Mellon, Lockheed Martin, and the Eglin and Wright Patterson Air Force bases, have been developing prototype interfaces and intelligent-agent coordination algorithms for interacting with small teams of Wide Area Search Munitions, or WASMs. WASMs are a cross between unmanned aerial vehicles and munitions. One of the major challenges, Lewis says, is trying to control multiple independent WASMs in close proximity to one another as they search for targets to destroy. Among the solutions: more elaborate communications systems capable of managing large WASM teams.
“We have to find ways to command them cooperatively,” Lewis says of WASMs.
The academics of nonstate violence
One doesn’t need to look much beyond the so-called Arab Spring uprisings throughout North Africa and the Middle East, or the chaos of Somalia, or the insurgent Taliban, or Mexico’s drug culture to realize that national preparedness and security in the future will depend largely on a new and better understanding of the world’s fast-emerging threats. That is what GSPIA Professor Phil Williams calls such violent nonstate actors as terrorists, criminals, insurgents, pirates, militias, warlords, and drug traffickers, among other armed groups.
Williams, the Wesley W. Posvar Chair in International Security Studies and director of GSPIA’s Matthew B. Ridgway Center for International Security Studies, has redirected the center’s research focus to target this threat category—and develop an academic program around it, in collaboration with the likes of the Carlysle, Pa.-based U.S. Army War College’s Strategic Studies Institute. He has found plenty to observe.
“If you look around the world, there has been a phenomenal rise of violent, nonstate actors,” says Williams, who singles out Mexico and the Middle East to make his point. “Around the world, something is going on with states, where we have more weak states, unstable states, and even failed states. And because of globalization, the instability is spilling over to developed countries. So we have a much less stable world where things are much less predictable.”
Williams says the Ridgway Center has been developing an academic effort “to get a handle on the new threat.” The new endeavor also has led to the addition of several new security studies faculty members in GSPIA, which has increased the Ridgway Center’s depth. “We’re building a niche within that subject,” he adds.
Among Williams’ own academic interests, which focus largely on the “pernicious” interconnectedness of these armed groups, are the following: the relationship between terrorists, criminals, and drug-trafficking organizations; the question of whether drug organizations in Central and South America are willing to help fund terrorists; links between those drug-trafficking organizations and terrorist groups in West Africa; and, more academically, the question of whether these threats represent disparate threats that seem to be converging to create fewer but bigger threats—or threats that are diverging, creating more threats about which to worry.
Williams also notes that, in cyberspace, the tools developed by criminals for cybercrime also are being employed by states as part of their own cyberwar strategies—yet another focus of Williams’ national security-related research.
As Williams, in defining the future academic challenges of the Ridgway Center, told an audience during his inaugural lecture, “The rise of violent, nonstate actors is one of many developments that have made the security agenda in the 21st century both more crowded and more complex.”
His contention, which drives his research and teaching, is, “The issue is not simply that governance is inadequate to meet security challenges; rather, it is that poor governance itself is a major source of insecurity and disorder.”
The emergence of an academic discipline
Recognizing a need to turn all of this collective research into a big-picture academic discipline with substance, Balaban and Sochats have spearheaded a Center for National Preparedness-based initiative to develop an entire University curriculum around national preparedness. As such, the team, in partnership with Pitt’s College of General Studies, launched in Fall 2010 an 18-credit Certificate in National Preparedness and Homeland Security program.
The program focuses on the analytical and managerial aspects of preparedness at the international, national, state, and local levels. It describes itself as evidence-based, analytical, systems-of-systems-oriented, and visual.
“We’re trying to develop scholars in this area,” Balaban says of the certificate program. “We now have the vision of what a Pitt graduate in this area should look like: one who adopts an analytic, often innovative approach, to managing issues of national preparedness.”
As the courses largely reflect what the collective research at Pitt has shown over time, Balaban and Sochats do acknowledge that their own views and approaches to preparedness have evolved, extending to their academic pursuits. Initially, for instance, the two subscribed to an “all-hazards” approach to disaster prevention, response, management, and mitigation. That is, they focused more on the various actors responding to a disaster and the commonality of response elements that can be identified across a variety of disaster scenarios. Recently, and based on extensive research, the team has adopted an “all-needs” philosophy, which means focusing more on the immediate and longer-term needs of the population that is affected by the disaster.
A problem with the all-hazards approach is that “the effectiveness of response cannot be measured exclusively in terms of logistical indicators, but rather by the recovery and resiliency of the region,” according to a paper by Balaban, Sochats, and Potomac Institute researchers Donald Donahue and Stephen Cunnion that was published in February 2012 in Homeland Security Affairs.
The article also states, “By basing planning on the needs of the impacted population—the ‘all needs’ approach—planners can better prioritize the full range of requirements and fully integrate both the government and nongovernment contributions.”
The Center for National Preparedness’ new certificate program builds on that philosophy, Balaban and Sochats note.
Says Balaban: “We’ve learned it well enough that we actually can distill it and teach it, and we’re focusing on developing the leaders of tomorrow.”