IEEE Electrification - December 2022 - 9

optimize power flow, detect grid
anomalies, deliver more reliable
energy at a lower cost, and unlock
opportunities for end users to
become active participants in energy
applications.
A software-defined smart grid
chip (Figure 3), developed by Utilidata
in collaboration with Nvidia, will
power next-generation edge intelligence
to increase grid resiliency,
decarbonization, and consumer
value. The smart grid chip is an open
source platform that collects and
analyzes grid edge data in real time,
using local rendering of machine
learning algorithms to identify and
Figure 3. The Utilidata smart grid chip powered
by the NVIDIA Jetson edge AI platform.
classify problems, quickly resolve
issues before they evolve into outages,
enable greater efficiency through
local operational
decisions, and seamlessly
connect and
manage DERs. GPUaccelerated
parallel
processing can solve
stochastic optimization
problems with
time-varying constraints
and targets,
enabling the management
of tens of
millions of DERs in
real time while improving the reliability
and resilience of the entire grid.
Let's imagine a scenario where a
storm took out a major part of the
grid. Control centers lost communication
connections to many parts of
the grid, and repairing the centrally
controlled network would take considerable
time. If, however, a sufficient
number of DERs remained
functional and could be managed
locally by edge intelligent devices,
these DERs could be used to power
the significant number of customers
during the recovery time. Edge AI
will ultimately shift the operational
Edge AI will ultimately
shift the operational
paradigm of power
systems from
centralized to truly
distributed, softwaredefined,
and
autonomous systems.
paradigm of power systems from
centralized to truly distributed, software-defined,
and autonomous
systems. An AI edge
platform is also an
ecosystem, which
will host and manage
applications
from many providers
to tackle different
energy management
issues. Eventually,
the power grid edge
will be transformed
into a lively marketplace
for customers
to shop for services and service providers
(including utilities) to create
more value streams for customers.
The Future of AI and
the Electric Grid
Digitizing the grid edge with open
source AI is a key enabler to accelerate
decarbonization and better serve
utility customers. Distributed intelligence,
powered by AI and edge computing,
helps
operational ize
real-time data from every point on
the distribution grid. To learn more,
see Utilidata's pilot smart grid chip
deployment (Figure 4) with Lake
Placid Municipal Electric (https://utili
data.com/lake-placid-smart-grid-chip
-pilot-project/), and find out how
Nvidia AI is powering a more sustainable
future (https://www.nvidia.
com/en-us/industries/energy/power
-utilities/?ncid=progr-679351-vt31#cid=
_progr_en-us).
Biographies
Yingchen (Y.C.) Zhang (yzhang@
utilidata.com) is the Vice President of
Product Solutions at Utilidata Inc.,
Providence, RI 02903 USA.
Marc Spieler (mspieler@nvidia.
com) is the head of global energy
business development at Nvidia,
Houston, TX 77096 USA.
Figure 4. Edge intelligence deployment.
IEEE Electrification Magazine / DECEMBER 2022
9

https://utilidata.com/lake-placid-smart-grid-chip-pilot-project/ https://utilidata.com/lake-placid-smart-grid-chip-pilot-project/ https://utilidata.com/lake-placid-smart-grid-chip-pilot-project/ https://www.nvidia.com/en-us/industries/energy/power-utilities/?ncid=progr-679351-vt31#cid=_progr_en-us https://www.nvidia.com/en-us/industries/energy/power-utilities/?ncid=progr-679351-vt31#cid=_progr_en-us https://www.nvidia.com/en-us/industries/energy/power-utilities/?ncid=progr-679351-vt31#cid=_progr_en-us https://www.nvidia.com/en-us/industries/energy/power-utilities/?ncid=progr-679351-vt31#cid=_progr_en-us

IEEE Electrification - December 2022

Table of Contents for the Digital Edition of IEEE Electrification - December 2022