THE ENERGY INDUSTRY TIMES - NOVEMBER 2018
13
Industry Perspective
A
rticial intelligence (AI) is
one of a suite of disruptive
technologies that promises to
transform our world.
Strip away all the techie-speak and
AI becomes a very relatable concept,
particularly if we apply it to some-
thing tangible, like an autonomous
vehicle. A camera, embedded in the
vehicle, detects the lines on the road;
it uses logic to assess whether to turn
left or right; and it initiates the ac-
tion. It sounds plausible. But what
can AI do in a sector that is all about
pipes, wires, grids, call centres and
customers?
In fact, the three principles that
can be applied to autonomous vehi-
cles, work for the power and energy
sector too. As a technology, AI does
three things: identies patterns; ap-
plies logic and initiates an action.
AI is so-called because it incorpo-
rates an element of reasoning typi-
cally associated with living things. It
enables tasks – usually repetitive, la-
bour-intensive tasks – to be per-
formed much more rapidly and accu-
rately than a human being could ever
do. So, in the power and utilities sec-
tor, it could feasibly sift through
masses of data to identify patterns; it
could apply logic that determines
how to respond to anomalies and ini-
tiate the appropriate response. Far
from replacing human ingenuity, it
complements it.
Right now, there are three principle
uses for AI in the power and utilities
industry. These are:
n Efciency savings. EY recently
undertook research that predicts that
Europe and Australia have just three
years until non-utility solar and
battery systems reach cost and
performance parity with grid-
delivered energy. Between 2023 and
2025, electric vehicles (EVs) should
achieve price and performance parity
with combustion engines. And we
have a decade or more before it
becomes cheaper to generate and
store electricity locally than to
transport and distribute it. While
these projections are subject to
geographic variations, two things are
certain: (1) once these tipping points
are reached, we will change how we
produce, distribute and use energy
forever, and (2) time is running out.
AI can help make existing ways of
working more efcient, reducing
costs and resources that could be
better deployed in the energy transi-
tion. Many of the traditional ways of
working are, indeed, ripe for an ad-
vanced technology intervention.
We might, for instance, use AI to
empower chatbots in call centres, so
that the rst few steps of customer
contact are fully automated, without
compromising the experience of the
end-user.
We could employ AI’s “deep
learning” capabilities – an articial
neural network that analyses differ-
ent layers of information to make
better predictions about the mainte-
nance of network assets, so that in-
tervention is timely but targeted. An
AI solution that can identify, with
99 per cent certainty, when an over-
head line warrants manual interven-
tion, will generate signicant ef-
ciency savings.
Or we could use AI to identify pat-
terns of behaviour that indicate cus-
tomer dissatisfaction, enabling inter-
vention and remediation to reduce
churn.
n Enabling the energy transition.
We are fast approaching the point at
which energy is neither created nor
consumed centrally. Consumers that
produce their own energy –
“prosumers” – will connect their
distributed energy resources to the
grid, and “prosumption” will be
dictated by variables such as weather
conditions and household needs.
Consumers will also connect their
devices – including smart appliances
– to the internet. Acceleration in
technology take-up means AI can
hive off data, pinpoint patterns of
behaviour and make predictions on
energy usage with greater accuracy
in order to deliver an intelligent,
stable and autonomous grid.
AI algorithms will, for instance,
recognise patterns of behaviour on,
say, a weekday evening in 2023,
when millions of EV drivers arrive
home and recharge their vehicles. By
distinguishing between drivers who
habitually use their cars overnight,
and those who leave it charging until
the following morning, the intelli-
gent grid will ensure that the battery
is sufciently charged in time for the
driver’s next journey, without exert-
ing simultaneous load on the grid
where possible.
n Accessing new revenue streams.
AI also provides an opportunity for
power and utilities companies to
access new business models and
revenue streams that will help them
to remain relevant beyond the energy
transition. They could, for instance,
use AI to compress, analyse and
monetise the huge swathes of data
moving through the energy
ecosystem, or follow the lead of
technology start-ups by harnessing
apps and other innovations to enhance
the networked and connected home.
Though AI articially enhances ca-
pabilities, many of its limitations are
the result of human trepidation.
For example, deep-learning AI al-
gorithms train themselves by sifting
through large volumes of data, and
from this they learn to identify ex-
ceptions to the norm and to make re-
liable predictions. Utilities therefore
need to ensure that they take steps to
structure and evaluate the data be-
fore introducing AI. If they do not,
there is a risk that the technology
will be ready while the data is still
being prepared. Ultimately, better
data produces a better AI learning
experience and improved outcomes.
Then there are issues relating to
computer power – or rather the lack
of it. Some utilities ght shy of mi-
grating to cloud computing solu-
tions, due to fears over data privacy
and cost. It is, however, all but a pre-
requisite for AI, given the technolo-
gy’s extensive storage and process-
ing needs.
Utilities also have to get to grips
with data privacy. They need to un-
derstand who owns the data; which
data is condential; and how open
data should be used and stored, if
they are to optimise its potential and
comply with relevant regulations.
There are exceptions. Some utili-
ties recognise that training a deep-
learning network takes dedicated in-
put and collaboration from both the
IT function and the business itself.
Increasingly, at EY, there are engi-
neers, shop oor workers, asset man-
agers and programme managers
working together on their AI capa-
bilities. They jointly dene and test a
use case, and populate the system
with relevant data – rather than
draining the entire data pool – to de-
liver the right algorithm training.
AI is a big data game. At EY, we
are working with organisations to
dene their data architecture, data
management and data governance.
By better understanding ownership
of the data and how it can be shared
and combined, meaningful algo-
rithms can be developed to underpin
trusted AI programs.
To make the most of AI’s potential,
boundaries are coming down – and
not just between IT functions and
other parts of the business.
While some utilities incubate their
own AI solutions in isolation, EY is
increasingly seeing evidence of a
growing tendency for businesses to
collaborate with other players – and,
in particular, existing start-ups.
EY has also seen utilities collabo-
rate with start-ups to access special-
ist capabilities – primarily in Germa-
ny, the UK, the US and the Middle
East. Notably, many are working
with omni-channel, intelligent cus-
tomer support applications, which
are essentially AI-powered chat solu-
tions that understand customer con-
versations and automate repetitive
processes – thereby reducing re-
source needs and costs.
Some start-ups even offer a plat-
form architecture for storing, con-
suming and selling energy, while
others work with utilities to deliver
predictive maintenance solutions.
By reducing unnecessary system in-
tervention, they enable timely reme-
dial action too, ensuring costs and
resources are focused in all the right
places.
I would go as far as to say that
collaboration or partnership is a
must for any utility. Otherwise, they
could struggle with the level of
technology sophistication and spe-
cialisation that more nimble start-
ups readily achieve.
So how far can AI go? Frankly, it’s
slow off the mark for some utilities;
others show varying degrees of AI
maturity.
Time is pressing. While there is no
need to invest huge sums right now,
start-ups will begin to erode utilities’
business models by developing AI-
enabled solutions that are smart –
and which customers like.
In conjunction with the Internet of
Things (which offers a virtual envi-
ronment through which distributed
energy resources can be connected)
and blockchain (which facilitates
trusted transactions between buyers
and sellers of home-grown electrons,
without the intervention of a central
authority) – AI has the potential to
reinvent energy delivery. Mean-
while, quantum computing – which
is still some way off but attracting
lots of investment – could be the big
game changer for AI. It will make
deep-learning networks faster, more
powerful and able to solve the tricki-
est challenges, all while storing even
larger bodies of data.
But even before all of these tech-
nologies reach maturity, utilities that
are not riding the wave of technolo-
gy innovation now risk losing some
or all of their business to competi-
tors. So, if they are to push ahead
with AI, they need to:
n Dene their AI strategy – an
absolute must
n Engage their business around how
they are going to achieve AI
transformation
n Start experimenting with AI as
early as possible, either by working
with start-ups or through in-house
innovation and acceleration
endeavours
n Run pilots and test cases to
understand what AI is and what it
could do for their business.
AI has a sixth sense, which en-
ables people to do things smarter
and eliminate repetitive tasks, in
turn reducing costs and improving
efciency. Of course, questions re-
main around how far AI can go in
the power and utilities sector, and
its long-term impact on human re-
sources. But those businesses that
take the initiative to start adapting
and testing the technology now will
certainly gain the competitive edge.
Thierry Mortier is Global Power &
Utilities Innovation Leader at EY.
The views reected in this article are
the views of the author and do not
necessarily reect the views of the
global EY organisation or its
member rms.
On the cusp of the transition in how we create, distribute and consume energy, Thierry Mortier, assesses the
practical applications of articial intelligence in the power and utilities industry.
Mortier: AI is a big data game. At EY, we are working
with organisations to dene their data architecture, data
management and data governance
Intelligent energy
transformation