pwr_april-2024 - 34

TECHNOLOGY
at a time when utility companies have
reported skyrocketing costs for liability
insurance, along with a decline in the
number of insurers and the addition of
built-in wildfire exclusions. Meanwhile,
on the labor front, companies are coming
to grips with the fact that they must
invest significantly in recruitment and
training efforts. This has to happen along
with offering competitive compensation
packages and advancement prospects to
attract and retain workers.
A Key Technological Advancement
Steps In: Computer Vision
As such risks and challenges increase,
the good news is that new technological
advancements have also stepped in
to help solve them. Recognizing the dire
need for new solutions, the utility industry
is increasingly investing in new technology.
Executives see this as a way to
assess and mitigate some of the utility's
major risks.
One prime area of innovation is ripe
with applications for the utility industry:
the field of computer vision. As a key
part of the fast-growing AI field, computer
vision involves building systems
that quickly obtain, analyze, and learn
from information gathered from images,
videos, or multi-dimensional data. It is
the technology at the core of facial recognition,
for example-the ability of your
phone to take a photo, " see " a pattern of
pixels, recognize that pattern as one that
matches your own unique face, and then
unlock your phone.
How does it work? Imagine a photo
taken of a power line in a forest (Figure
1). When humans look at the photo,
our eyes take in shapes, colors, and
countless other details, and send this
information to our brains. Our brains
have evolved over millennia to instantly
recognize patterns, turning them into
meaning without having to think about
it. We look at the photo and see every
tree, branch, bush, and the power line.
We might even focus in on a particular
tree branch to see if looks to be hanging
low to the power line.
For a computer, though, the photo begins
as just a series of pixels-assorted
strings of data. To " see " an image, computer
vision relies on machine learning
and neural networks. Machine learning
(or deep learning) involves building algorithms
and models that allow a computer
to teach itself to recognize and see what
it needs to see.
Over time, the network's predictions-that
is, the prediction that a cer34
2.
Example of a solar panel field image captured by a drone and analyzed with computer vision
to detect potential anomalies (captured in red and yellow). Courtesy: Genium
tain shape corresponds with an object it
wants to identify-become increasingly
accurate. Massive amounts of data are
needed for this kind of learning, however.
The computer must be fed image
after image of forests and power lines
in order for it to eventually be able to
distinguish between a power line and
a tree (let alone to " see " and assess
small details like a single branch's proximity
to a line).
These feats, while powerful, were not
widely available in the commercial arena
until recently. Roadblocks to their adoption
included the computing power needed
to store and analyze the vast quantity
of data collected from video and images,
as well as the need for engineers with
the expertise to create, train, and deploy
the machine learning models and neural
networks involved.
However, thanks to cloud-related cost
savings, the technology is now entirely
cost-effective. Rather than hosting the
supercomputers themselves, companies
can partner with cloud-based computing
services to store their data, along with
taking advantage of an increasing number
of pre-built deep-learning vision models
and platforms available from providers.
Furthermore, computer vision's costeffectiveness
increases profoundly
when its applications can mitigate the
costs of the challenges it is helping to
solve. We're seeing this now in the utility
industry in a number of applications,
including those described below.
Computer Vision in the Utility
Industry: Drones That Can See
Computer vision can be harnessed
through various means in the utility industry.
It's already being used in interesting
ways, especially when combined
with other rapidly advancing technologies,
such as autonomous robots and unmanned
aircraft systems-that is, drones.
One example is the deployment of
www.powermag.com
drones to aid with nuclear power inspections.
Nuclear utilities have used drones
to access and inspect areas where
safety may be a risk for a human inspector,
or that are otherwise hard to reach.
Drones can inspect construction activities,
concrete structures, hard-to-reach
components, storage systems, radiation
mapping, and more.
Computer vision enters the picture to
analyze the images collected; otherwise,
a human would have to manually review
every minute of footage collected. One
technology company developed machine
learning models to assist utilities in automatically
detecting defects in their concrete
containment structures. Using this
tool, nuclear utilities can deploy drones
to conduct inspections of the structures,
upload their image datasets, and then
use machine learning models to evaluate
and review the results.
Another example of the technology's
use can be found in the solar power industry.
For instance, Genium, a Silicon
Valley-based software and engineering
service company, has developed a solution
called SolarScan. It combines data
collection, processing, and analysis to
streamline and improve the solar panel
inspection process.
For example, one large European utility
company expanded its renewable
energy sources to include a large solar
field in Africa. In order to achieve optimal
performance at these solar panel
assets, the field must be regularly inspected.
Many kinds of obstructions
(such as sand, dirt, bird droppings, and
debris) can prevent optimal performance
or even cause lasting damage.
However, such inspections are laborand
time-intensive, all of which is complicated
by their geographic distance
from the parent company.
Genium's solution involves deploying
drones, such as Parrot's Anafi model,
to collect RGB (red-green-blue) and raPOWER
| April 2024

http://www.powermag.com

pwr_april-2024

Table of Contents for the Digital Edition of pwr_april-2024