A team of University of Miami industrial engineering students is helping to keep power flowing to the people.
With overgrown vegetation encroaching on electrical lines being one of the primary causes of power outages in the state, the students have partnered with Florida Power and Light (FPL) on a project that will help the utility company conduct more strategic and effective tree-trimming measures—and it all starts with remote sensing technology.
Aerial drones equipped with Light Detection and Ranging (LiDAR) devices create 3D digital twins of overhead infrastructure and the environment, allowing FPL to see where trees are growing too close to power lines and where trimming is needed most.
But such data, which is collected annually and covers thousands of miles of power lines, only reveals where vegetation is growing today, not where it will grow tomorrow.
"So, that's where we come in," said Gabriella Bueno, a senior industrial engineering major in the University's College of Engineering. "As part of our capstone project, FPL provided us with multiple years of LiDAR data and tasked us with analyzing it to create models that predict vegetation growth rates in the future. And with that data, FPL can plan maintenance earlier and concentrate their tree-trimming efforts to areas that need it most."
So far, Bueno and the other members of her team—Abril Orejon Salas and Braydon McCullough—have tested their predictive models using LiDAR data on vegetation growth rates from only two years. But as data from other years is incorporated and tested, their predictive models will become more accurate, allowing FPL to become more proactive in tree-trimming maintenance.
Current vegetation trimming has traditionally followed set three- and six-year cycles. But with predictive modeling of vegetation growth rates, the long-term goal, Bueno explained, is for FPL to shift from cycle-based trimming to an approach that targets high-growth, high-risk areas sooner.
"Fast growing species like palms or bamboo don't always align neatly with a fixed trimming cycle," said Erin Schreck, vegetation operations lead for FPL's east region. "Some areas could benefit from attention sooner, while others may not pose the same level of risk."
To improve the accuracy of their predictive models, the students have also examined how rainfall and other environmental factors can influence vegetation growth rates.
"These students are the analytical engine behind the work," said Tiffani Rodriguez, lead project manager in operations services at Juno Beach, Florida-based NextEra Energy. "They're working with real FPL data and tackling real operational challenges."
For Bueno and her industrial engineering team, the semester-long project brought classroom concepts to life.
"While learning about modeling and data analysis in class is incredibly useful, being able to apply what we've learned to a hands-on, real-world project has taken that learning to another level," Bueno said. "Working alongside a team like FPL's, who are so knowledgeable, supportive, and willing to guide us through every step, has made this project even more educational."
