Over the last decade, digital technology has transformed our grid infrastructure, helping utility companies strengthen power grids’ security, efficiency, and resilience. From enabling smart grids and installing reclosers to using LiDAR data captured by drones or aircraft to expedite manual inspection processes, companies have taken massive steps to digitise their operations in response to changing demands.
As innovations in artificial intelligence (AI) and machine learning (ML) continue to impact the smart grid, another technology is starting to play a pivotal role in enhancing powerline reliability: satellite-powered intelligence and ML algorithms to detect potential hazards.
By Dobrina Laleva, UP42
(R)Evolution of Satellite Technology
The technology revolution that has already transformed our day-to-day lives is also quietly transforming space. Satellites are no longer the huge constructions of the past – they have shrunk dramatically and are often no larger than a microwave. They also cost a fraction of what they used to and can be mass-produced. This has led to an explosion in data availability, coupled with advances in cloud computing and AI that enable the analysis of larger volumes of data.
In terms of numbers, there are more launches than ever, driving the cost per measurement further down. 2022, for example, saw 155 launches, 149 of which were successful. In 2022, 4,852 out of 8261 total satellites were active – an increase of 11.84% from 2021. 1030 of those satellites were used for Earth observation, second only to communications . Some of the many advances in satellite technology include the availability of very high resolution satellite data (think 0.3 m resolution), great revisit rates (multiple times a day), and a variety of images types available (infrared, multispectral, hyperspectral, SAR, and stereo imaging capabilities), among others. Satellite technology can also be used to strengthen the safety and security processes needed to keep power grids free from unpredictable faults and unwanted interference.
Augmenting Utility Vegetation Management
From roads and railways to power lines and pipelines, encroaching vegetation causes downtime, damage and catastrophes.
Utility vegetation management remains a huge line item in most annual budgets, involving manual inspection of thousands of infrastructure miles. But it’s mainly reactive maintenance, leaving very little focus on prevention.
There are many factors to be considered when considering a new approach to vegetation management and definitively great benefits in adopting satellite technology.
First, it provides a unique opportunity to augment existing methods, offering more scalability, reliability and flexibility. Satellites can capture long corridor lengths in a single pass and do so frequently. This approach drives efficiency and savings and removes the constraints of LiDAR and UAV photogrammetry. Utility operators can identify risk areas for further investigation with helicopter optical imagery, LiDAR and UAV photogrammetry to avoid costly image capture of the entire infrastructure. Or they can task a satellite for a specific time and cloud coverage, which improves reliability and predictability, reducing risks from bad weather. Furthermore, using satellites for the most inaccessible points of infrastructure maximizes compliance and minimizes the risk of danger to employees.
Choosing the Right Data
Satellites offer different types of useful data: optical satellites, for example, take imagery in the visible or near-visible portion of the electromagnetic spectrum using existing radiation.
When optical satellites take imagery of the same point from two or three different locations, we combine those images to construct digital surface models, which are important for utility vegetation management.
Synthetic Aperture Radar (SAR) sensors, on the other hand, capture imagery by transmitting microwaves at a target and catching the backscattered signals at any time, day or night. SAR’s microwaves can penetrate clouds, haze and dust, as well as dense forest canopies to give information on different layers of vegetation. Elevation data can be used to track differences in elevation (height or volume) and ensure that hazards such as vegetation encroachment are monitored, predicted, and mitigated, minimizing damage.
While access to archive elevation data is great, it is mainly used to establish a base reference. This is because existing data may cover only some of the Areas of Interest (AOIs) needed for a project; terrain changes have occurred as a result of construction or natural events, or more precision and control are needed over the data used to produce elevation models.
In such cases, companies can task a satellite to get fresh stereo imagery. Stereo pairs (overlaying two images of the same area taken from different angles) or tri-stereo pairs (overlaying three images for even more information) provide an accurate solution for building elevation models.
Stereo images enable the generation of 3-dimensional models, and naturally, tri-stereo images create more accurate 3-dimensional models than basic stereo. Satellites can capture image pairs either in one pass, along the same orbit (along-track stereo), or from different orbits (across-track stereo). Stereo imagery can be used to extract stereo DEMs and DSMs. Stereo imagery can also be complemented with LiDAR, SAR and survey data to extract insights from different datasets and get a better overview in more densely vegetated areas.
A Successful Story
A great example of rethinking vegetation management with satellites comes from an AI-first SaaS company, AiDash.
UP42, a geospatial developer platform and marketplace, helped AiDash transform the process for its customer, a Fortune 500 utility company that wanted to improve vegetation management along lines spanning over 80,000 km.
AiDash’s client relied on a traditional approach, with trimming cycles spaced across 4–5 years and manual data collection. This time-consuming process gave them minimal predictive capabilities for future outages, damages and costs.
To solve the problem, AiDash developed a proprietary AI model called the Intelligent Vegetation Management System (IVMS), with satellite imagery from UP42. Through the UP42 platform, AiDash accessed both archive and tasked very high resolution (50cm) Airbus Pléiades imagery of utility lines, providing detailed coverage for the customer’s remote asset monitoring needs.
AiDash was able to predict species’ growth rates and height levels compared to individual utility towers. The company prepared a 3- to 5-year trim plan for the entire network with an accuracy of over 85%, improving system reliability by 15% and reducing its annual budget for vegetation management by 20%.
There are countless more potential applications for utility vegetation management – from combining weather data with multispectral data to forecast strong winds affecting trees, to analyzing the state of surrounding forests for risk of fires.
With the recent emergence of dedicated geospatial platforms offering easy access to complex data, satellite imagery has proven to be the complementary tool of choice by many utility operators and solution developers, enabling more targeted and efficient solutions.
ABOUT THE AUTHOR
Dobrina Laleva is a senior product marketing manager at UP42 with over 10 years of experience in business development and product marketing for highly technical products.
At UP42, Dobrina leads full-cycle go-to-market strategies for new features, products, verticals and partnerships.