Vegetation Condition Assesssment by Multispectral Imagery

Risk Reduction Category

Environmental Monitoring

Technology Description

Vegetation threat analysis refers to identifying vegetation that may encroach upon powerlines, transformers, and support structures. Depending on line voltage, elevation, and local concerns, vegetation clearance requirements may range from as little as 1 foot (0.3 m) for a low-voltage conductor to 13 feet (3.0 m) for a high-voltage line at high elevation [7]. Vegetation should be considered when it grows towards utility assets (e.g. grow-in threat), and when the vegetation extends above assets and could fall on a conductor or other component (e.g. fall-in threat).[6] Trees stressed by drought leads to additional risk of premature falling into power lines. Additionally, areas impacted by drought may have increased amounts of dry biomass ‘fuel’ underneath the tree canopy.

Utility rights of way are often inspected manually. Even with the aid of helicopter or fixed wing aircraft, any imagery that is collected during the inspection is analyzed post-flight by humans. Further, utilities routinely dispatch arborists to manually inspect trees on a watch list or to place additional trees on the watch list – a time consuming and costly exercise.

Because utility Vegetation Management (VM) programs are one of the largest recurring maintenance expenses for electric utility companies in North America [5], it is important to identify ways of reducing these costs. Technology improvements including automated UAS, multispectral imagery, and image processing offer new opportunities for utilities to improve the efficiency and accuracy of their vegetation inspections. Multispectral cameras can extend the information gathered during aerial inspections, capturing information that is otherwise indetectable with standard visual (RGB) cameras. Most multi-spectral cameras capture RGB imagery along with imagery in both the infrared and ultraviolet wavelengths. The extended bands can be captured at the same frame rates and resolutions as their visual counterparts, providing layers of imagery that can be analyzed by humans or AI.

  • Visual (RGB) – captures bands detectable with the human eye (red, blue, and green)
  • Infrared (IR)– detects heat signatures. This band is ideally suited for vegetation condition because chlorophyll, an indicator of healthy vegetation, reflects in the IR band.
  • Ultraviolet (UV)– can reveal arcing, partial discharge, and flame. While perhaps not the primary focus for vegetation inspection, this band can detect arcing caused by overgrown vegetation. Because it is included in the data being gathered, it can be simply another layer of intelligence gathered on the ROW.

What is envisioned is a periodically self-deploying UAS, located in a sheltered charging station, that would fly a programmed (or self-guided) route along utility rights of way. Imagery collected by the UAS could be uploaded to a cloud service or processed locally. Machine learning would analyze the imagery, detect conditions of interest, and send alerts upon detection of these anomalies. This same equipment can be dispatched for various other purposes such as asset inspections, fault location, or even suspected fire ignitions.

Technical Readiness (Commercial Availability)

The vision for the future of aerial vegetation inspections involves a complex combination of state-of-the-art technologies:

  • Automated UAS flight - As of 2023, most UAS deployed by utilities are manually operated, however, as technology improves automation is increasingly possible. A challenge with automated aerial asset inspections has to do with the ability to capture useful imagery [1]. Lighting, truncation, and obscuring of the subject can be factors. This challenge is being addressed by improvements in navigation technology. Another of the immediate challenges today involves FAA regulations against flying a drone beyond line of sight of an operator. Gradually, the FAA is loosening some of these restrictions and waivers are being granted within specified conditions.
  • Camera technology - As of 2023, multispectral cameras are suitable for drone mounting, having portability and capability to capture images from a moving camera at focal distances of a few meters. Multispectral cameras are ideally suited and commonly used for vegetation health assessment in a variety of ecological use cases. Purpose-built RedEdge cameras focus more narrowly and specifically on the band centered on the wavelength reflected by chlorophyll.
  • Data processing and anomaly detection – As of 2023, with sufficient training data, machine learning has been demonstrated to be effective at detecting anomalies in imagery. Vegetation assessment may involve detecting trees that are leaning or fallen (detectable with RGB bands) and assessment of health of vegetation (detectable with the infrared band).

Combining these technologies into a vegetation inspection use case is the next step toward the vision of fully automated vegetation inspections. More needs to be known about the efficacy of the data gathered, the ability to retrieve actionable information, number of false positives and missed positives, and costs of the effort.

The following list of manufacturers is the product of an Internet search using a general description of the technology as the search term. Sometimes more than one variation on the search term is used. The objective is to identify the most demonstration-ready products available in the category. Toward assessing demonstration readiness, the manufacturer websites typically provide useful information such as writeups of successful use cases or field demonstrations, number of deployments, or other indicators. Where lack of information exists online, further inquiry is made by phone. Generally, one to three frontrunners emerge as being most ready for a field demonstration. Preference is given to manufacturers who sell to the United States, or, if emerging technology, those who have participated in US-based field demonstrations.

Implementations / Deployments

While the technologies involved in carrying out the use case are commonly used today, combining them effectively into a vegetation condition use case is the next step. More needs to be known about the efficacy of the data gathered, the ability to retrieve actionable information, number of false positives and missed positives, and costs of the effort.

Innovations as of Mid 2023

Potential Enrichment Work Opportunity

References

[1] UNMANNED AIRCRAFT SYSTEMS (UAS): Advanced Payloads. EPRI, Palo Alto, CA: 2018. 3002015063.

[2] Weisenfeld, Neil et al. “Infrared Scanning Reveals Defects.” Tdworld.com. https://www.tdworld.com/underground- tampd/infrared-scanning-reveals-defects. (Accessed September 2018).

[3] Rebecca DelPapa Moreira Scafutto et al. “Evaluation of thermal infrared hyperspectral imagery for the detection of onshore methane plumes: Significance for hydrocarbon exploration and monitoring.” International Journal of Applied Earth Observation and Geoinformation 64 (2018) 311-325. https://www.sciencedirect.com/science/article/pii/S0303243417301411. (Accessed September 2018).

[4] Santovasi, Steve and Evans, Laron. “UAS Technology Offers Soaring Potential.” Burnsmcd.com. https://www.burnsmcd.com/insightsnews/tech/uas-technology-offers-soaring-potential. (Accessed September 2018).

[5] Federal Energy Regulatory Commission. “Tree Trimming & Vegetation Management.” Federal Energy Regulatory Commission. https://www.ferc.gov/industries/electric/indus-act/reliability/vegetation-mgt.asp.

[6] Maximizing the Value of Right-of-Way (ROW) Unmanned Aircraft Systems (UAS) Collected Data: Four Applications for Remotely Sensed Data. EPRI, Palo Alto, CA: 2020. 3002018898.

[7] Vegetation Indexes for Hazard Tree Management. EPRI. Palo Alto, CA. May 2020. 3002017448. https://www.epri.com/research/programs/025032/results/3002017448