IEEE Power & Energy Magazine - May/June 2022 - 4

Ning Lu and Steve Widergren
a long road to singularity
making inroads with artificial intelligence
D
4
DID YOU HEAR THE ONE ABOUT
the self-driving car that wrote its autobiography?
I hear the groans, but introducing
an issue devoted to artificial
intelligence (AI) and machine learning
(ML) needs an icebreaker. IEEE
Power & Energy Magazine topics tend
to focus on power system challenges
and then the cool techniques or technologies
being applied to address them.
With the topics of AI and ML, this issue
inverts the paradigm to explore the
data science and methodology underlying
AI as a disciplinary toolbox for
power system applications.
As professionals and members of
the digital age, we cannot avoid benefiting
from AI in our daily lives and
being exasperated by its
shortcomings.
Examples include search engines,
digital assistants interpreting speech,
call- and web-based chat services, and
autonomous, intelligent controls in all
sorts of devices and systems.
AI is beginning to penetrate power
system operations and planning as
well. It is a complex field full of jargon
such as feature selection, hidden layers,
learning rates, and test and training
data. Although a single magazine
issue cannot make one an expert in AI
and ML,
the contributors to this
issue
provide insights that peel back the
cover to explain a bit of how AI works,
where techniques are being adapted to
power system applications, and where
it struggles to gain trust.
Digital Object Identifier 10.1109/MPE.2022.3150815
Date of current version: 19 April 2022
ieee power & energy magazine
In This Issue
In the past decade, we have witnessed the
rise of AI. Thanks to the rapid advancement
in data science, computing hardware,
and the theory of computation, researchers
have made a quantum leap in
training AI applications to play complex
board games, process images, and perform
a wide range of natural language
processing tasks such as translation,
summarization, classification,
and extraction.
In recent years, with one
success after another,
AI-powered software
has demonstrated its
exceptional potential
in social media, entertainment,
health care,
education, finance, surveillance,
robotics, and
transportation. The list
grows longer each year.
Encouraged by the
ach ievement of AI
programs with performance
comparable or
superior to their human
counterparts in data-intensive
tasks, researchers
in many domains
are rethinking standard techniques for
data processing, information extraction,
model development, and automation.
Optimists believe that the fourth technological
revolution, fueled by AI-infused
systems (i.e, systems with AI incorporated
into the decision-making process), is
underway. If true, the breadth and depth
of upcoming changes will profoundly
transform our society, and our way of
living, working, and entertaining.
However, thus far, when managing
As professionals
and members
of the digital
age, we
cannot avoid
benefiting from
AI in our daily
lives and being
exasperated
by its
shortcomings.
and operating critical infrastructures,
the acceptance of AI remains limited
to applications in low-risk, low-impact,
labor-intensive, decision-support
systems. For example, AI adoption in
power grid operation has been limited
to planning and predictive applications
such as load forecasting,
anomaly detection,
secur ity assessment ,
synthetic data generation,
and system model
identification. This is
because the critical nature
of secure and reliable
operation of power
grids demands trust
in
decision making in all
situations, whether previously
seen or completely
unexpected.
Integrating AI into the
decision-making process
requires human operators
to understand the reasons
behind AI-generated decisions.
Awareness of the
consequence and risk of
each AI-driven action needs experiential
support to engender faith in the technology.
Establishing trust between human
operators and AI-powered applications
requires proof of interpretability, generalization,
and robustness of their decisions.
To introduce the status of AI applications
in the power industry, Guest
Editors Tao Hong and Fangxing Li
may/june 2022
editors' voice
IEEE Power & Energy Magazine - May/June 2022

Table of Contents for the Digital Edition of IEEE Power & Energy Magazine - May/June 2022
