Beyond the Trim Cycle: How Pointerra Helped WISPR Build a Smarter Approach to Vegetation Management
Written by: Greg Itzstein, COO at Pointerra
Vegetation-caused outages are the leading cause of power interruptions in the US Northeast. Utilities conventionally rely on fixed trim cycles, trimming every circuit on schedule regardless of actual risk. Today’s digital tools make that broad approach unnecessary for such a precise problem.
Trim cycles made sense when utilities had limited data and few alternatives. They make less sense now, when high-resolution lidar can map every tree near every span, outage records can train models to predict failures, and optimization tools can show which tree removal cuts outage risk the most per dollar. With the data and models available, WISPR set out to test whether they could be combined into a single decision framework utilities could use.
The early answer is yes — and the results show real potential. Broader validation across more circuits and geographies will sharpen the framework further.
What is the Role of WISPR?
WISPR (the Wind Impact Study for Power Resilience) is a DOE GRACI-funded consortium that ran from June 2024 to October 2025. Pointerra served as program lead, our partner utilities provided circuit access and operational data across Massachusetts, New York, Maine, and Connecticut. Three university teams contributed the framework building firepower: Cornell University on physics-based tree risk modeling, the University at Albany on wind projections, and Colorado State University and the University of Connecticut on machine learning outage modeling and economic optimization.
The $1.63 million DOE grant was initiated by our utility partners and submitted by Pointerra, with a deliberate design principle baked in from the start: no single entity could build this alone. The utilities had the operational data and real-world constraints. The universities had the modeling rigor and peer-reviewed credibility. Pointerra had the high-resolution 3D data platform to connect them. The consortium structure was not incidental — it was the methodology.
The Four Pillars of WISPR in a Single Framework
WISPR integrated four technology components that utilities typically hold in separate systems, and the integration is what makes the results meaningful.
High-Resolution LiDAR and Digital Twin
Vehicle-mounted mobile laser scanning was deployed across 10 circuits, capturing centimeter-level 3D point clouds of every tree, pole, wire, and structure at road speed. Individual trees were segmented and attributed with height, crown area, diameter at breast height, and proximity to conductors — the precise geometric inputs that Cornell's biomechanical failure model needed to compute failure probability under wind loading. The data was hosted in the Pointerra3D platform on AWS, accessible to all consortium partners as a common data environment.
Importantly, the digital twin built for WISPR is not a single-purpose asset. The same data collection that fed the vegetation risk models also enables clearance analysis, pole lean and sag trending, attachment spacing audits, and joint use surveys. Once the data is collected, it becomes a multi-use infrastructure intelligence platform. For utilities thinking about how to justify the investment, the parallel with storm response is direct — the same digital twin that supports pre-storm vegetation risk prioritization also powers post-storm damage assessment and restoration.
Historical Outage Intelligence
A decade of span-level outage records (2014 to 2024) trained a transformer-based machine learning model to learn the relationships between wind speed, wind direction, vegetation density, seasonal patterns, and outage likelihood. The model achieved an AUC of 0.783, representing strong discrimination between outage and non-outage conditions. The key insight from this component was directional: sustained high winds, not gusts, are the primary outage driver, and wind arriving from the 260–360° and 0–20° range showed the highest frequency of associated outages across the study circuits.
Wind Projections at Circuit Level
The University at Albany applied Weather Research & Forecasting (WRF) dynamic downscaling and a Thermodynamic Global Warming framework to project wind conditions at circuit level through to mid-century. The projections are granular enough to assign directional risk scores at the span level — identifying which trees and spans are most exposed to dominant wind vectors rather than treating all vegetation near a line as equally hazardous. High-density circuits with more than 500 trees per mile showed the strongest correlation with projected outage risk, while low-density circuits showed minimal incremental risk — an important finding for investment prioritization.
Economic Optimization
The optimization framework, developed by Cornell and Colorado State, is where risk scores become investment decisions. A budget approximation algorithm allocates spending across spans by marginal resilience return per dollar, comparing three mitigation strategies: tree cutting at approximately $400 per tree, pole elevation at $30,000–$70,000 per span, and undergrounding at $50,000–$200,000 per span. Life-cycle cost analysis using present-value calculations allows fair comparison across a multi-decade horizon.
The output is a ranked investment roadmap defining how and where resilience is optimized for every dollar spent.
Three Findings That Reframe the Vegetation Management Problem
- Precision beats proximity. Vegetation risk is not simply about how close a tree is to a line. It depends on tree geometry, the direction of wind loading, the biomechanical failure mode — uprooting at lower wind speeds versus stem breakage at higher ones — and the geometric relationship between the tree's fall arc and the conductor. When LiDAR data, outage records, GIS circuit topology, and unit costs are unified in a single framework, span-level risk can be calculated rather than estimated. That is a meaningful shift in how utilities can justify and target their vegetation management spend.
- Risk concentrates on a handful of spans. Across every circuit studied — across diverse geographies, vegetation types, population densities, and customer profiles — the vast majority of outage risk came from a small number of spans. Crews do not need to walk every line. They need to know exactly which spans to target. That knowledge is what converts field time and inspection budgets from broad coverage into precision deployment.
- Investment returns peak early. The biggest resilience gains came at modest investment levels, typically $100,000 to $200,000 per circuit. Most circuits showed a pronounced inflection point around $500,000, beyond which marginal returns dropped sharply. The framework also identifies not just which spans are at risk, but which mitigation action delivers the best return at that specific location — tree cutting, pole raising, or the more capital-intensive option of undergrounding, which was rarely justified within the wind-hazard scope alone but may be warranted when wildfire, ice, and snow risk are added to the model in future iterations.
The circuit-level ROI figures illustrate the range: Circuit A returned $2.42 million in avoided outage value on a $200,000 investment. Circuit G returned $11.38 million on $50,000. Low-density circuits in the study showed negative ROI, confirming that fixed-cycle trimming is not warranted everywhere — which is itself a valuable finding for utilities managing O&M budgets under regulatory pressure.
What This Means for Vegetation Management Practice
WISPR demonstrated that the industry can consider a new framework to move from schedule-driven to risk-driven vegetation management. The shift requires integrating data that utilities already hold — outage records, GIS circuit topology, customer impact data — with high-resolution LiDAR geometry and physics-based modeling that most utilities have not yet applied at scale.
The consortium model matters here. As our university partners noted during the NARUC webinar presentation, real-world constraints including trees on private property, state and federal land access, environmental permits, and community approvals are as important to the decision framework as the risk models themselves. Academic modeling rigor without operational context produces findings that cannot be implemented. Operational expertise without modeling rigor produces investment decisions that cannot be defended to regulators. The two need each other.
Our utility partner’s regulatory context is instructive for the broader industry. New York and Massachusetts both require defensible, evidence-based mitigation strategies in climate resilience action plans. A risk framework like WISPR — with peer-reviewed methodology, span-level granularity, and economic optimization outputs — has the potential to support the kind of investment justification those filings require.
Open Questions Worth Exploring Through WISPR
WISPR was designed as a study, not a final answer, and the consortium was deliberate about naming the limitations. Only three mitigation strategies were modeled, yet others can be considered. Cost data was derived from literature rather than utility-specific accounting. Model transferability will vary by local vegetation species, terrain, and maintenance practices. The optimization framework focused on vegetation-hazard strategies — mitigation approaches that address wind, ice, and wildfire risk simultaneously, may shift the economics of undergrounding significantly over a 30 to 50-year horizon.
The regulatory questions the study raises are equally open. At what combined risk threshold does undergrounding become the efficient long-term choice? How should risk ownership be allocated when the highest-risk tree sits on private property? And as risk probabilities become calculable rather than estimated, how do utilities, regulators, and communities work together to translate that precision into investment standards?
These are questions for the industry as a whole, not for any single utility or program. WISPR provides a framework for answering them with data.
Want to Learn More?
Watch the Full Presentation
The WISPR consortium presented the full study findings at the NARUC Innovation Webinar Series in May 2026, including the lidar visualizations, circuit-level ROI analysis, and the academic modeling methodology. The recording is available on the NARUC’s website and YouTube Channel.
View it directly here on YouTube.
Get in Touch
If you'd like to discuss how this framework could apply to your network, we'd welcome the conversation. Schedule a Demo to see our vegetation management tools in action for various customer use cases.
Download the Whitepaper
For utility vegetation managers wanting to understand how Pointerra3D operationalizes this data across the full UVM lifecycle — from classification to work planning to change detection — our vegetation management whitepaper goes deeper.
