Laura McLay: Crunching data on high-risk scenarios
Laura McLay, a new associate professor of industrial and systems engineering, is an expert in operations research, which she defines as “the discipline of applying advanced analytical methods to help make better decisions.”
Laura McLay’s research canvas is massive data — banks of millions of emergency 911 calls, commercial airline flights and ship cargo deliveries — which she uses to help improve decision-making in these high-stakes endeavors.
As a data challenge, it might seem like searching for the proverbial needle in the haystack, but McLay is quick to clarify her goal. “You actually never find the needle,” she says. “You just make a better haystack.”
McLay, a new associate professor of industrial and systems engineering at UW–Madison, is an expert in operations research, which she defines as “the discipline of applying advanced analytical methods to help make better decisions.” While not yet a household word among academic fields, operations research received a major awareness boost after the Sept. 11 attacks, as an important tool in improving airline screening and security and combating terrorist threats.
McLay works to develop mathematical models to identify how to design risk-based passenger-screening methods, given that the “haystack” of data can be used as a type of passenger-risk assessment tool. “You’re really trying to weed out all of the low-risk events and be able to focus on maybe the 5 percent you really need to worry about,” she says. “A lot of times operations research is exactly the right tool to use, because it answers the question: How do we best utilize limited resources and imperfect data?”
While not yet a household word among academic fields, operations research received a major awareness boost after the Sept. 11 attacks, as an important tool in improving airline screening and security and combating terrorist threats.
As a former professor at Virginia Commonwealth University, McLay found a robust new research target after the 2010 “Snowmageddon,” a storm that dumped more than three feet of snow on parts of the Washington, D.C. region. In the aftermath, there were controversies over poorly managed emergency resources, response times and decisions.
This was a research perfect storm for McLay, who married her longtime work in analysis of emergency response times with data on severe weather. Her analysis of the Snowmageddon response, compared against more typical weather days, revealed important differences that could improve decision-making and prioritizing calls during future weather emergencies.
Emergency responders develop notoriously good instincts in reducing their response times, but McLay hopes her work bolsters those instincts with hard data on managing calls during high-volume or high-risk periods.
“When the system is overwhelmed, some kind of triage or priority reclassification makes the most sense, even when it goes against the nature of responding to everything in the same way,” she says. “There is an expectation with public services that on the worst day of your life — like a weather disaster — emergency people will be there for you. The reality is, there are a lot of factors that might not make them available, at least not immediately.”
“You’re really trying to weed out all of the low-risk events and be able to focus on maybe the 5 percent you really need to worry about.”
One of McLay’s primary focus areas for emergency response times is in calls involving cardiac arrest, where literally seconds can make a life-or-death difference to the patient. This happens to have a strong weather connection as well, since the highest levels of cardiac arrest occur during periods of heavy snow.
McLay does much of her work in partnership with municipalities, given that there is so much variation at the local level, such as urban vs. rural settings, or full-time vs. volunteer crews. She is in the process of establishing some of those local connections in Wisconsin this year.
Speaking of connections, McLay has managed to connect operations research to punk rock, in the form of her public interest blog “Punk Rock OR.” Since she does so much work in the public domain, McLay wanted a title that reflected the field’s beneficial aspirations. “And punk is sort of socially aware and wanting to make the world a better place, so it seemed like a good fit,” she says.
In the blog, McLay delves into the seemingly limitless opportunities to apply operations research to daily life, including marathon running, winning the lottery, planning your wedding, predicting NCAA tournament outcomes, or fitting three child seats in a Honda Civic (McLay is a mother of three).
Occasionally her posts will go viral, such as the one about stochastic processes.
“Actually it was on vampires and stochastic processes,” she adds. McLay has always been struck by the stability of vampire populations in most vampire films, since branching process models would dictate vampires having exponential population growth.
“I find the zombie movies much more mathematically consistent than vampire movies,” she says, and on that front offers a helpful blog post on how to optimally prepare for a zombie apocalypse.
“People like to talk about how some of these mathematical principles apply to everyday life,” McLay says. “It’s a lot of fun and a great way to give back to the field.”