Tree Intelligence: FPL controls ever-growing plant life

One of the most beautiful but fragile landscapes in the continental United States, Florida is a difficult landscape for power utilities to manage because it combines dense vegetation with an often extreme climate that includes hurricanes, frequent thunderstorms and the downed trees that these severe weather events bring with them. For Florida Power & Light Company, using advances in technology to manage these demands is critical to delivering superior reliability to its customers.

FPL service territory includes more than 85,000 miles of power lines in 43 counties spanning two time zones with many species of trees and vegetation. Perhaps the most daunting challenge facing FPL is Florida’s year-round growing season. In the Sunshine State, there is enough light, warmth and humidity to ensure that plants never stop growing. This may require more frequent data collection to correctly capture changing conditions. FPL is also working to ensure it can use systems and data for hurricane response, leveraging digital twin technology against post-storm analytics to help with post-storm damage.

The more detailed data FPL can obtain about its infrastructure, including vegetation that may affect it, the better it is able to serve the utility’s 5.7 million customer accounts that support more than 11 million residents across Florida. Many utility vegetation programs rely heavily on visual inspections and rely heavily on ground crews for vegetation management. Utilities without visual intelligence typically prefer fixed vegetation cutting cycles and rely on pen and paper to manage their schedule. With a digital solution, they can make better, more informed decisions about where and when to prune vegetation.

Some utilities are open to changing their inspection processes and survey data portfolio as long as the new options allow them to maintain or improve reliability at a lower cost. Over the past decade, LiDAR collection costs have come down so utilities can accurately identify trees and risks and develop annual preventative and reactive work plans using analysis and AI.

FPL began using LiDAR for transmission lines in 2016 and expanded it to distribution facilities through a series of proof-of-concepts in late 2020 and limited pilots in early 2021. The utility is working currently expanding its use statewide.

FPL has partnered with GE Digital to develop a platform that integrates GIS, LiDAR and analytics to create a 3D digital twin and 2D GIS-style visuals and reports. The team also integrated the results into existing work management electronic workflows to optimize field resources.

Prior to working with GE Digital, FPL began transforming its program from largely relying on paper forms and field notes to a fully electronic workflow. The new process allows jobs to be scoped, scheduled and handed over to a tree crew electronically. LiDAR laid the foundation for future improvements in remote sensing. The process also allows FPL to use LiDAR and other data sources (RGB, hyperspectral, multispectral, FLIR, etc.) to refine and automate its scope of work, which will further improve team efficiency. and will save inspection time. The system will also be interoperable with Electric Office, Smallworld and GE’s ADMS, reducing the time, complexity and errors of data transfer and integration. The utility can now make better decisions about where to cut, which ultimately benefits customers.

GE’s AI-based vegetation management system, called Visual Intelligence, has two components: data capture and vegetation management analysis. The potential value generated by using visual images based on data captured by many different sources (drones, fixed-wing aircraft, helicopters, satellites, humans, sensors) to guide vegetation management decisions is based on this analytical approach:

  • First, moving from an annual rotational cut strategy to a purely risk-based approach that identifies and categorizes the risk of excessive encroachment across the entire region. This approach leads to the highest priority cut of the current year and recognizes the uniqueness of the service area.
  • And second, move from a distance-based vendor payment model to a volume-based payment model. This approach shifts payments from miles/kilometres driven to acres actually cut.

For FPL, the most important advantage of the system is the ability to work in three dimensions rather than two. FPL has historically scheduled work based on linear miles or frames, but not all miles are equal. LiDAR data allows vegetation program managers to be more accurate and efficient in assigning work.

Visual intelligence takes a data-driven approach to vegetation management, asset inspection, and asset inventory using visual intelligence. Inspection data can be collected using a variety of means, including drones, fixed-wing aircraft, helicopters, satellite data, as well as extensive visual inspections. This data can be collected over a wide service area that will benefit some major electric utilities around the world.

AI-based classification for LiDAR point clouds and photogrammetry enables automated classification of vegetation and trees in the area of ​​interest. Machine learning automates vegetation growth calculations and 2D/3D dangerous tree maps, then produces risk-based cutting plans as shapefiles specifying cutting prioritization, volume calculation cutting edge and ground clearance.

Automated analysis workflows allow personnel to focus on critical issues that require their immediate attention, spending less time and energy reviewing visual inspection datasets. By integrating visual inspection data with the recording system, overlaying the data with the geographic information system (GIS), the utility gains essential context.

Using a data-driven vegetation management approach is an important factor in mitigating damage to network facilities and avoiding weather-related outages and fires. GE Digital’s Visual Intelligence Suite also allows utilities to deploy asset inspections in their territory to aid in storm recovery, enabling economies of scale and data coordination.

FPL and GE Digital have also worked together in the area of ​​microgrids. In 2021, FPL opened a Microgrid Control Lab at the University of Central Florida in Orlando, which serves as a research center and offers computer science and engineering students the opportunity to work in a utility control room.

Exploring and leveraging new technologies to continually improve customer service is part of FPL’s culture. So, as the utility works to operationalize LiDAR for vegetation management, it is also mindful of how this technology can be used and adapted elsewhere. This includes working to link data to asset management systems and make it available for a wide variety of utility inspections.

James V. Hayes