Panelists included Justin Kramer, director of emerging technologies and data analytics for the Orlando Utilities Commission; John Matelski, executive director for the Center for Digital Government*; and Joshua Sheldon, GIS director for the East Central Florida Regional Planning Council.
Key takeaways from the discussion included the following:
- Get your data in the right place: “I think for us, it’s a lot of forecasting, predictive analytics, different considerations in that space. So, I think it starts with getting your data right. Building models on top of that allows you to see how extreme weather events are changing, and that’s a good way to understand how we’re reacting to those weather events,” Kramer said. For example, “what were the outage predictions and considerations from last time? How are we improving those? And then it even gets into better predicting how the generation and load are in a storm event.”
- Lean into machine learning: “We have been doing AI machine learning for what, 50 to 60 years?” Matelski said. “Machine learning models analyze historical disaster data to predict likely future events. That’s some of the modeling that’s been done. Then, of course, on the cyber side, there’s cyber-attack pattern recognition; those types of things have been getting better and better as the technology has evolved over the years.” He recommended using predictive simulations for disaster scenarios and what-if analyses to help inform preparedness plans.
- Don’t just plan for one scenario: “We’re in a time right now where there’s so much data, and there’s so much modeling going on,” Sheldon said. “We always say not one model is correct, but a lot of models are really useful. So, you always have to go through everything and not just plan for one scenario, but you want to have the whole gamut of everything, because when a disaster actually happens, you wonder if it’s going to happen on the ground.” For example, “a tree can fall and direct flood waters somewhere else. So, it’s really just kind of being prepared for anything that could happen.”
- Meet with different departments: “What we used to do for part of our planning was go out and meet with all the departments and agencies and find out what their needs were, what tools they had available and what tools they wanted us to look for,” Matelski said. For some departments, specific tech uses included geospatial data or geographic information systems, Internet of Things devices, social media monitoring and cyber incident monitoring tools.
- Understand the citizens you serve: “One of the things that we’re working on a lot is looking at these hazards, but also really trying to understand the people in the communities in our region,” Sheldon said. “You care about where those hazards are and where those events might occur, but you also care about who you’re going to be serving if a disaster happens, like understanding that maybe for more of the population, English isn’t their first language, or if they rely on public transit. A lot of times, our most vulnerable population gets hit by these hazards the hardest, and they don’t have the resources to bounce back as easily, so being able to focus recovery and mitigation efforts to make sure that these hazards don’t hit them as hard is important.”
