The Bionic Network Company, #4: Building a Data-Driven Vegetation Management System

Wildfires have become a growing concern worldwide. Utility networks play an important role in mitigating this impact of climate change by minimizing the risk of fires due to vegetation contacting the lines.

Vegetation management is not a new focus for networks. Given the amount of spend (tens or hundreds of millions of dollars, the largest operating expenditure for most networks), the level of exposure to risk (including financial risk), and all the advances in technology and analytics in this space, it’s clear that vegetation management has been a top priority for many networks. But there’s still room for improvement. 

Recent advances in predictive data science provide networks with the opportunity to bolster vegetation management performance considerably. But embracing the latest technology available is not enough.  As the first article in our bionic network series explains, companies need to take what BCG calls a “bionic approach,” an advanced digital capability that combines the best of what digital and human have to offer. Bionic networks with advanced vegetation management have reduced costs from 20-30% without increasing network risk, significantly uplifted employee engagement and performance, and led much more informed discussions with customers and regulators. 

What Bionic Networks Do Differently 

For the past decade or so, companies have been introducing various technologies, such a light detection and ranging (LiDAR) or high-resolution satellite imagery, to capture data on the current state of vegetation and determine whether it posed a risk to power lines. Bionic networks have taken this approach one step further. They use predictive analytics to foresee future risks, identify targeted actions to mitigate them, and reduce the need for manual inspections and audits. 

Specifically, bionic networks do four things differently: 

(1) Start with Clear and Ambitious Business Outcomes 

They start by focusing on the business outcomes they want to achieve rather than the latest features of a specific technology. Bionic networks are clear on how the new process leveraging data and analytics can help deliver benefits and augment their way of working. 

• Vegetation management spend focused on highest risk span and treatment methods. The aim is to reduce risks in areas with the greatest potential impact.
• Lower inspection and audit costs. Automated inspections and audits result in optimal frequency and higher quality data.
• Optimal cutting cycles. Growth prediction and cutting optimization will reduce the volume of trees that are cut and the frequency with which they are cut without increasing the risk that the vegetation will grow too close to the lines.
• Contractors deployed surgically. Networks tell the contractors exactly when and where to cut or remove trees, thus eliminating unnecessary travel, scoping, and treatment costs.
• Better prices from contractors. Networks help their contractors become more efficient (doing better resource planning, scoping and planning work remotely, avoiding duplicative audits, and minimizing travel); in turn, contractors lower their prices.
• Contractors' performance tracked and measured digitally. Transparent and traceable data in close to real-time allow optimal monitoring of contractors’ performance.
• More efficient engagement with regulator. Advanced vegetation management is key to engage with the regulators based on outcomes rather (i.e. risk) than inputs (i.e. frequency, distances), which is a key prerequisite to break the trade-offs and achieve better outcomes for customers. 

(2) Focus on Delivering Value Fast 

Bionic networks use agile ways of working, with cross-functional teams building predictive models in numerous test-and-learn iterations to develop a minimum viable product. Agile ways of working are usually more efficient and effective than the traditional approach.  

One utility we worked with developed a lidar-based predictive model and rolled it out with field contractors after the first six months. The team then kept refining the model, improving the digital interface, and adding functionality – but they already captured most of the savings in year one. 

Another client introduced a tool to capture efficiencies in the short term while working on end-to-end optimization and predictive modeling in the medium term. The solution is based on field image capture an AI-enabled app that allows rapid tagging of hazardous trees and that feeds into the backlogs of the field. Employees, contractors, and even customers can take pictures of the hazardous trees. The app collects key data and AI-based analytics assigns a criticality score to be backlogged and ranked for work execution. 

(3) Take an End-to-End Journey (E2E) Perspective

It’s critical to design the target state with the E2E process in mind, rather than a specific piece of technology since the solutions in this space are evolving very quickly. For some networks, granular field-based data combined with publicly available data on species and weather has proved the most efficient approach; others have preferred LiDAR or aerial photography. Yet another option is multi-spectral high-resolution satellite stereoscopic imagery, which is reaching maturity. 

Instead of waiting for the best technology—and missing hundreds of millions of dollars in savings opportunities—bionic networks start by mapping out how the process should work to achieve the goals they are looking for. They then determine which technologies for data capture and analytics can best deliver the insights needed to measure the current state of vegetation, predict vegetation growth and health, define treatment priorities, and optimize treatment in space and time. The audit performed at the end of the process provides data that are used to finetune the strategy, creating a continuous feedback cycle. 

(4) Make Contracting Relationships More Collaborative 

Strong contractor collaboration, including joint planning and bilateral data sharing, is critical to enable improved vegetation management. Bionic networks are going further, creating strategic, long-term relationships and collaborating with contractors on a day-to-day basis. They are also providing incentives for contractors to refine predictive models to provide more precise (that is, smaller) estimates of the volumes of trees in need of treatment. To offset the smaller work volumes, networks are also introducing targeted volume reduction programs. 

We are also starting to see contractors deploy predictive analytics in the negotiation process, to be better informed about the future work volumes. Networks that don’t have the same level of knowledge as their contractors are leaving the money on the negotiation table. And networks that have the knowledge but don’t share it with their contractors take the risk that the contractors will bake additional resources in their work estimates to be on the safe side. 

Consequently, it’s really important for networks to rethink the way they engage, collaborate, and share data with contractors so that the new process is a win for both sides. Networks that have better visibility can offer contractors a more precise view on volume of work at better prices, reducing spend on contractors by 20% to 30%. 

*** 

We at BCG believe that utility vegetation management is about to undergo massive and exciting change. The latest technological progress opens a range of improvement opportunities. Networks that want to create impact from the beginning of change should start with the business outcome in mind and iteratively implement solutions adapted to their needs with an E2E perspective.

About the authors:

• Oxana Dankova, Managing Director and Partner
• Alfonso Abella, Managing Director and Partner
• Konark Singh, Managing Director and Partner
• Carolyn Ford, Managing Director and Partner
• Joseph Butler, Partner
• Hanna Schumacher, Project Leader
• Santiago Gallego, Associate Director

This is the fourth in a series of BCG articles on the bionic network company. Please find here the first three: "Introducing the Bionic Network of the Future", "How to Get More from Workforce Management Systems", and "Developing Digital Capabilities to optimize Asset Management and Performance".