Mechanical Engineering Professor David Morton.
In 1999, just eight years after the collapse of the Soviet Union, American and Russian officials were still trying to account for and secure the large stockpile of nuclear weapons, and nuclear material, leftover from the USSR's 69-year regime.
The stockpile — which at its peak was believed to total more than 45,000 nuclear weapons — had been haphazardly guarded after the regime's collapse, and officials feared that nuclear material could be smuggled out of Russia.
What they needed were nuclear detectors that could be stationed along Russian borders. More than that, they needed insight on where to position the detectors along the borders so they could account for vulnerable exit points — ones where smugglers would be most likely to cross or have higher chances of doing so without being caught.
At The University of Texas at Austin, three mechanical engineering professors and their students had developed a model that could provide such insight. The optimization model helps predict behavior of nuclear smugglers who are smart, have multiple entry and exit points and want to avoid being caught.
Reliance on sophisticated models like this one has grown in recent years, and so with it has the field of operations research and industrial engineering (OR/IE). Those outside of the discipline, which uses advanced analytical methods to help make better decisions, have long misunderstood it. But through organizations like IBM, BNSF and Microsoft, the analytics field has experienced a rebirth. OR/IE is used by everyone from major airlines to dynamically set the price of airline seats and by major league baseball teams who use analytics and statistics to recruit the best baseball players during a draft, as portrayed in the book and movie "Moneyball."
The nuclear detection model, developed by mechanical engineering Professors David Morton, Elmira Popova and Erich Schneider is just one example of how faculty at The University of Texas at Austin are using OR/IE research to tackle problems around the world — from providing insight on how to prepare for a flu pandemic, to respond to global climate change and to organize work shifts so that hospitals can more effectively treat patients.
"What I love about working in this field is that we can apply the same mathematical tools to a range of challenges, and in each case we gain a better understanding about the nature of the optimal solutions," Morton said.
In 1999, Morton and his research team used a comprehensive database, known as PATRIOT and operated by Los Alamos National Laboratory, to provide government officials with a priority list of where to station the detectors along borders in Russia, Kazakhstan and Ukraine. A few years later, their model was used again to boost security during the 2004 Athens Olympics.
"One school of thought is to just put detectors at the busiest border crossing where the most traffic goes through, but that's assuming the smuggler would follow the flow of traffic and take on more risk," Morton said. "Using an OR/IE approach, our model looks at vulnerabilities in the network. Because if you put detectors at four busy crossings and leave two less busy crossings open, there's a chance a smuggler will take advantage of that."
Staying ahead of a growing, emerging field
Morton and other faculty and students in the Cockrell School of Engineering and the McCombs School of Business hosted roughly 4,500 analytic experts from around the world for the Institute for Operations Research and the Management Sciences annual meeting. The meeting, held in 2010, was the highest attended in the organization's history and also marked the first time The University of Texas at Austin hosted the event. The university’s selection is a testament to its small but strong OR/IE program.
"We've hired exceptionally talented faculty over the years, and because we're a small group we can't limit ourselves to just one special skill so we must be diverse in all areas," said Jonathan Bard, a professor in the Department of Mechanical Engineering and program coordinator of the university's OR/IE program.
This year the program partnered with local startups in manufacturing, inventory management and software development to create a semester-long practicum for students to work on industry projects.
"We've developed close research partnerships with industry and our students are gaining hands-on research opportunities that lead to them to highly competitive jobs," Bard said.
Bard’s research has paired him with manufacturing and electronic assembly companies like Texas Instruments, which uses OR/IE to improve its planning and scheduling of semiconductor fabrication.
As an example of the diverse applications of the field, Bard recently worked with the M.D. Anderson Cancer Center to develop a more efficient daily schedule for its 200 nurses. The schedule accounts for times of day when the hospital is typically most busy as well as the fact that most nurses work 8-12 hour shifts, three to five days a week.
"It's a very vexing problem because hospitals run 24 hours a day and they need coverage the whole time but demand varies over the day," Bard said.
He recently won a grant with The University of Texas Health Science Center at San Antonio to better match interns and residents with clinic assignments.
Offering insight in face of uncertainty
The greatest aspect of OR/IE is its ability to shed insight on research issues with great uncertainty.
Eric Bickel, an operations research professor in the Department of Mechanical Engineering, has spent the past few years tackling what he calls the "mother of all decision-making-under-uncertainty challenges," — climate change.
"It's a huge capital investment with lots of uncertainty," Bickel said.
Using in-house analytical models to test his theory, Bickel recently released a major report suggesting that the effect of global warming could potentially be ameliorated by engineering ways to reflect more sunlight back into space — a form of intervention known as geoengineering.
The report co-authored by Hudson Institute Fellow Lee Lane was selected by a panel of international experts — which included four Nobel laureates — as one of 16 areas of research that governments and philanthropists should prioritize to respond to the world's most pressing challenges. It also ranked first among the four papers solicited by the experts on climate change, and 12th overall on a priority list released in late May by the Copenhagen Consensus Center.
"The main point of our paper is to ask whether research into climate engineering is justified, and our conclusion is that it is," Bickel said. "Climate engineering holds the potential to limit warming, and it's also the only technology that could possibly cool the Earth quickly enough if needed. So it plays an important risk management role."
Faculty in the university's OR/IE program have used their expertise to provide insight on everything from studying the probability of loss-of-coolant accidents — a mode of failure for nuclear reactors — to Morton's latest research on assessing and predicting the State of Texas' preparedness for an influenza pandemic.
Working with Biology Professor Lauren Ancel Meyers, Morton recently built optimization models that made use of statewide hospital discharge data from over the past 10 years. The models helped the researchers determine how many medical ventilators — which are needed to treat critically-ill patients with influenza — would be needed in Texas in the event of a flu pandemic.
"The H1N1 flu was a wake up call that we need to be better prepared in the event that there is a more serious flu pandemic," Morton said. "These can be devastating viruses, and many of the research projects OR/IE researchers tackle have high-stakes consequences. The hope is that with analytics we will not only better understand them, but we can mitigate their impact when they do occur."
