When a hurricane is dangerous enough to prompt evacuations, thousands of people find themselves fleeing at once.
Emergency planning officials want to know the best ways to safely and quickly evacuate their residents. That’s often meant focusing on a single objective, like moving people out of danger in the fastest way possible. But researchers at the FAMU-FSU College of Engineering and Florida State University’s Department of Psychology have developed models that account for multiple considerations in a crisis, including the physical and mental demands on evacuees, especially vulnerable populations.
Their work was published in the Journal of Management in Engineering.
What is the main conclusion of the paper, and how is that finding important?
Dulebenets: We looked at emergency evacuation planning for Broward County, a South Florida county where residents must sometimes evacuate because of hurricanes. We analyzed things like the average utilization of evacuation routes, utilization of emergency shelters and other criteria.
Our key conclusion was that emergency managers should look beyond criteria like total evacuation time or distance. They should consider things like the mental and physical demands placed on evacuees and the frustration they endure while preparing for and fleeing a hurricane. Increasing mental and physical demands can lead to more problems in an evacuation; for example, crashes along the evacuation routes. These criteria may influence the best way to assign evacuation routes and emergency shelters to people, especially in vulnerable populations.
Ozguven: In tackling evacuation issues, the importance and role of demographics and socioeconomics, or human factors in general, are often disregarded or given less importance. The most common goal of the emergency agencies is to evacuate people from the affected region in the least amount of time. However, vulnerable populations such as seniors might need special help evacuating. Thinking about their particular needs can allow for less stressful evacuations, and that ultimately saves lives.
This study was conducted before we were dealing with COVID-19, an issue that complicates an evacuation. For example, will we need more shelter locations with fewer people per shelter in order to mitigate crowding? Could the modeling account for that issue in some way?
Ozguven: Yes, the model can handle those changes. It is all about changing the inputs: How many people will need special assistance, whether it is related to the virus or some other disability? Which shelters will provide the extra capacity? These questions can be answered by incorporating sensitivity analyses, which help us understand how the model’s output is affected by different input parameters.
What were the various objectives you wanted to optimize in your study?
Dulebenets: The proposed mathematical model accounts for six goals. We want to minimize the total evacuation time of the people leaving, their total mental demand, their total physical demand, the amount of time it takes them to prepare, their total effort and the amount of frustration they’ll endure. Our goal is to be able to account for all this in a given metropolitan area where the impact of a natural disaster is expected to be catastrophic.
Based on the number of variables, solving a mathematical model for a real evacuation — say, one that involves 500,000 people — may take a prohibitively large time to compute, even for a supercomputer. So we focused on developing heuristics for multiple objectives that can quickly assign evacuees to the available evacuation routes and emergency shelters based on our criteria.
Why should emergency management planners consider things like frustration or effort required by evacuees?
Boot: It’s important to think about the human factors of evacuation, especially in Florida, where many of the evacuees may be older adults. That drive away from an evacuated area may be more challenging for older adults. We know that as people get older, there are perceptual and cognitive changes that can make driving more challenging in general. During an evacuation, those challenges may only be exacerbated because of the difficulties and stress.
How do you calculate something like “frustration?”
Ozguven: “Frustration” is hard to quantify, but we used surrogate measures to estimate how frustrated people were while evacuating. One of those was evacuation time. The longer it takes, the more frustrated people are. We also simulated evacuations in a driving simulator and asked study participants questions about how mentally and physically demanding various tasks were and how discouraged and stressed they felt during those tasks. Those answers and information about the demographics and socioeconomics of the participants provided our input information.
What lessons can emergency planners take from this research?
Ozguven: Recent hurricanes such as Irma and Michael showed that many vulnerable populations, like senior citizens, were reluctant to evacuate for a variety of reasons. They were afraid it would be dangerous to leave, or they were worried about their property or language barriers. Other people didn’t take action because they had special needs, physical disabilities, cognitive impairments, concern for pets or a lack of money. These problems have clearly been causes of senior deaths in these hurricanes. So emergency planners can use the models we developed that account for some of those issues.