May 20, 2025 • 13 mins
In this episode of Energy Transition Talks, CGI’s Peter Warren speaks with AI experts Diane Gutiw and Lukas Krappmann about how artificial intelligence is transforming the Energy & Utilities sector. They explore key insights from Hannover Messe, how AI is closing workforce gaps, optimizing hydrogen systems, and driving efficiency through cross-industry innovation. Discover how AI is enabling faster decisions, predictive maintenance, and real-time operations in a sector under increasing pressure to adapt and evolve.
Visit our Energy Transition Talks page
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Peter Warren (00:00):
Hello everyone and welcome back to part two of our
discussion on AI in energy andutilities how this is impacting
the energy market.
It's part of our ongoing energytransition conversation.
In part one, diane and Lucastalked about what a digital twin
is, digital triplets and howit's being applied.
Today, we're going to pick up afew more things on innovation
(00:20):
and the impact of it, but whydon't we start with a
reintroduction of yourselves,diane?
Do you want to go first?
Diane Gutiw (00:27):
Great, hi, peter.
Thanks for having me back.
My name is Diane Gucu.
I'm a vice president at CGI andI lead our global AI research
center.
Over to you, lukas.
Lukas Krappman (00:36):
Yeah, thanks, diane, for the introduction.
Yeah, so my name is LukasKruppmann.
I'm from Germany and thereforeresponsible for all of our
manufacturing, or more theHaltrim business there and
during the past.
We're up with a couple ofcompanies and trying to actually
generate more data from thevalue of the product.
Peter Warren (00:54):
Thanks very much and just bringing in some
information from our ongoingsurveys that we do with our
customers and other folks thatare not our customers in the
industry.
We call it our voice of theclient survey, so our voice of
the customer survey, I guess iscorrect.
We pointed out this year thatand this is not fully published
yet but there's a lot of peopleinterested in AI.
(01:16):
Obviously they're interested inautomation.
They're trying to do more withless.
They're trying to manage thefact that they're not being able
to get all the people they want.
They've got cutting budgets,they've got the fact that they
have people retiring withknowledge.
All of that is sort ofhappening all at the same time.
One of our customers said thiseverything's happening
(01:37):
everywhere all at once, and Ithink that's a quote from a
movie.
Perhaps you know, diane, what'syour thought on the most
exciting trends for energy andtransitions?
How do you see this AI actuallybeing able to fill those
problems I just described?
Diane Gutiw (01:51):
You know, I think we saw a lot in Hanover and I
know we all just were at theHanover Messe event and there is
so much going on.
It's hard to nail down onething Automation and the move to
agentic AI to extend what wewere already doing in
intelligent automation intoagentic AI is brilliant.
Also, the efficiencies thatorganizations are gaining
(02:15):
through use of some of thesegeneral purpose tools.
Peter Warren (02:17):
And Lucas.
I mean we also saw theconfluence between the logical
layer, which we just sort ofdescribed, but also the physical
layer of which we just sort ofdescribed, but also the physical
layer of devices like yourclients are using, but also the
overlap into physical security,just a ton of things, the use of
imagery, lidar, radar,overlapping and bringing really
(02:38):
a bigger holistic viewpoint topeople.
But what's your viewpoint onsort of what was the most
exciting thing that we picked upat Hanover and where you see
your customers going?
Lukas Krappman (02:46):
Yeah.
So, frank, what's interestingfrom, or what was interesting
from my perspective, was thatNow everything's basically
available in terms of technicalknowledge and the tools,
(03:20):
macroeconomical challenges,meaning how everything comes
together and how the wholeecosystem, for example, of
manufacturers, energy companies,logistic companies and also
health and life science, canreally work together to solve
the problems which are burningum on our plate right now.
Yeah, and this is, I think,especially due to the
geopolitical and other economicinfluences, really the thing
which is most interesting formost of our clients out there.
Peter Warren (03:42):
Yeah, you bring up a couple of good points.
There is that the supply chainsor the ecosystems are all under
attack.
People are having to re-vector.
What some people would take twoyears to do now has to happen
in like two days.
We've also heard from a voiceof the clients.
People are looking for new,better software tools to be more
agile in both their regulatoryand tariff type of management,
(04:04):
given what's going on.
So there's a lot of things thatpeople are trying to deal with.
Diane, how do you see AIplaying a role in all of that?
It seems very complex.
Diane Gutiw (04:14):
Yeah, it's a really good point, but when we come to
innovations, we're at a goodplace in having tools that are
allowing us to do things thatwere really complex, expensive,
complicated to do in the past.
You know, in addition to theeconomic climate globally, which
energy is absolutely at theforefront of, we also have a
rapidly aging populationglobally, and when Lucas is
(04:36):
talking about parallels acrossdifferent industries, that's a
challenge all industries aretrying to solve.
How can we do more, do morepersonalized, specialized type
work with less people being ableto do it?
And, as you mentioned earlieras well, a lot of the people
(04:57):
that have the deep knowledgeinto systems are retiring.
So, when we look at AI, some ofthese things that are the
challenge maintaining, retainingknowledge, upskilling people
quicker, being able to take oversome of the menial tasks so
people are working at the top oftheir skills rather than
spending days of their weeklooking for information,
collating information and tryingto do analysis.
(05:18):
There's tools that are notreplacing people, but are
replacing tasks, and replacingtasks that could be done much
quicker, perhaps moreefficiently, so that people are
able to do what they need to doto be creative.
I think across all industries,we're starting to see this
challenge and finding patternsand opportunities both to reduce
(05:39):
risk and improve efficiencies,and improve those opportunities.
Peter Warren (05:43):
Yeah, if I can ask an open-ended question to both
of you and you can let me knowwho would like to answer first,
the cross-industry thing.
We've hit it a couple times andwe're seeing at least from our
point of view as CGI, because welook across industry that the
problems we're trying to solvein one industry actually
parallel or have already beensolved in another industry.
And looking to the benefit ofthose things, whichever one of
(06:06):
you would like to go first,maybe you'd make a comment on
how we're actually learning fromother industries and pulling
things back in.
Diane Gutiw (06:12):
Sure, I'm happy to take that one because I think I
probably spend my time in themost different industries lately
.
I think there is bettercollaboration now.
Certainly you know ourselves atCGI.
We spend a lot of timesupporting the specialists
across the different industriesfor that very purpose.
It's a best practice in datascience as well is to not
(06:34):
reinvent the wheel, to findefficiencies and reuse patterns.
So a good example would bepredictive maintenance.
What we've done in drilling inthe past long predating
generative AI we've reused in anumber of different industries
to look at how to determinepotential faults in different
types of equipment when it'shard to nail it down.
(06:55):
So clustering that data andthen using a K-shape algorithm
to nail it down.
It doesn't matter if that'sjust mining or utilities or
anything that's using heavyassets or any assets you know,
even looking at modalities inhealthcare.
That type of model and approachto solving problems can be
shared.
So the population retiring isthe same thing.
(07:17):
Where are their efficienciesand where in a workflow, in a
value chain that we can enhancewhat we're doing with AI rather
than replace, so that the peoplecoming in are upskilled much
quicker.
People that are doing the workare able to focus on a
meaningful task rather thanmenial tasks.
That's a problem that we're alldealing with, and I think there
(07:37):
definitely are efficiencies.
Our vendor partners are good atsharing that across industries,
and I think that's one of thereal values that we have as our
team at CGI.
Peter Warren (07:46):
So, Lucas, at the Hanover Messe you ran a workshop
with customers on digital twinsand digital triplets.
What was sort of the keymessage that you took away from
that?
What were the clients lookingfor?
What were the clients lookingfor?
What were they interested inmost?
Lukas Krappman (07:59):
Yeah, I mean like it was a pretty interesting
workshop, to be honest.
So we've raised the workshoparound digital twins and
triplets in the hygieneecosystem, but the participants
were actually from a couple ofindustries.
So we had, for example,participants for the hydrogen
industry, we had manufacturers,but we also had participants,
(08:20):
for example, from agriculturalbackground, and the most
important thing all threebasically have in common is that
they kind of like have to makea lot of data available with
certain speed and and thengenerate insights from it.
So how do you connect, forexample, data coming from a
(08:42):
farmer's field, from crops, forexample, to an infrastructure
and add a business applicationon top so that you can do a
proper farming or, for example,put out water for the seeds if
there's rain in the weatherforecast, stuff like that?
This is, for example, for theagricultural perspective and
then for the energy andutilities industry.
(09:03):
For example, I had a couple ofcustomers with a more like a
hydrogen background and theirchallenge basically is on how
they can integrate the data from, for example, the hydrogen
electrolyzers then into, forexample, large energy grids, on
how to optimize productiondepending on the weather, do a
lot of predictive maintenancefor example, when do we have to
replace a particular stack.
(09:24):
Is it better, for example, tonowadays turn on my hydrogen
electrolyzer to generatehydrogen or the energy, or
should I look for like ananother source to do it?
Yeah, so it's actually aboutconnecting all the different
data points out there in thedifferent ecosystems and then
making an informed decisionbased upon it.
As you might know, there's alot of data out in the industry
(09:48):
and in the ecosystem.
There is a lot of metadata outof it.
That's even a more importantthing.
It's not only about thedifferent signals you have out
there, but also the metadata andhow you process it, and I think
this is going to be a realchallenge in the next couple of
years.
Peter Warren (10:06):
Lucas, I really appreciate what you said on that
because it really highlightsthe data and it was interesting
that a use case from what we'retalking about hydrogen was so
applicable to somebody in theagricultural industry.
And what they were trying to dowas an interesting dialogue
with the researchers and otherfolks that were in the session.
Just heading to wrap up hereand give you guys a chance to
sort of summarize what do youthink is the vision on the
(10:30):
global stage, what's the bigthing moving forward?
Where is this all going?
And maybe we'll go back to you,lucas, first and give you a
chance to give that summary andthen we'll toss it to Diane.
Lukas Krappman (10:40):
Yeah, I'm like when you were walking around
with all the different booths,right, so you saw that AI is
actually part of the thing.
Every booth.
I didn't see like one boothwhere AI wasn't incorporated or
incorporated into the businessmodel, right.
And this starts from, likeintelligent co-pilots, from
custom GPTs, for example,different use cases from energy
(11:02):
suppliers out there.
From my perspective, and whatwe talked about to different
customers.
It's really about how AI cancontribute to, for example, the
efficiency of each of ourcustomers out there, how their
day-to-day work actually can getmore efficient and bring more
benefit to them.
And with that, I guess I'llhead over to Diane.
Diane Gutiw (11:23):
Yeah, you know, I think the thing that's happened
in this last year is AI nowisn't just one thing, it's not
just generative AI that peoplehave on their phones.
I find that innovations aresplitting into three things, and
those three things are reallywhat we saw in Hanover, as Lucas
said, as well as what's goingto drive us forward.
The first is tools for internalefficiencies.
(11:45):
This is your operational,administrative access to
information, access to data tobe able to do things in your
day-to-day workflow quicker.
So a lot of what came out ofthe GPTs and software
development and being able tointegrate with everyday work
tools is going to be a hugeimpact on every sector, but will
(12:07):
be able to help a lot.
The second would be automation.
We already see a lot ofautomation in this sector, lots
of robotics and automation andprocess engineering being
advancing technologies in thisspace.
I think that by havingsolutions that are able to do
(12:28):
more, having more autonomousautomation, having automation
that's very focused on morecomplex tasks which were hard to
complete without really clearinstructions, you're able to do
more.
And then the last one is data.
We've talked about data a lot,but using AI to be able to mine
that data, to find patterns andhave more evidence-based answers
(12:50):
to questions.
You know, to me those threethings are splitting out of this
technology evolution that we'veseen and those are, I think,
what the three top areas thatare going to have a real
positive impact on the energysector.
Peter Warren (13:02):
Well, I'd like to thank you both, and I will add
one thought too is this was aconversation we had with other
clients.
Diane, when we were there, isusing generative AI to actually
error correct and check data andvalidate, so AI is actually
helping us improve our data inreal time.
So that's a great thing.
Well, thank you both, reallyappreciate this.
So hopefully everybody enjoyedthis session and we'll do more
(13:24):
coming up in the coming weeks.
Thank you, lucas.
Hello everyone and welcome back to part two of our
discussion on AI in energy andutilities how this is impacting
the energy market.
It's part of our ongoing energytransition conversation.
In part one, diane and Lucastalked about what a digital twin
is, digital triplets and howit's being applied.
Today, we're going to pick up afew more things on innovation
(00:20):
and the impact of it, but whydon't we start with a
reintroduction of yourselves,diane?
Do you want to go first?
Diane Gutiw (00:27):
Great, hi, peter.
Thanks for having me back.
My name is Diane Gucu.
I'm a vice president at CGI andI lead our global AI research
center.
Over to you, lukas.
Lukas Krappman (00:36):
Yeah, thanks, diane, for the introduction.
Yeah, so my name is LukasKruppmann.
I'm from Germany and thereforeresponsible for all of our
manufacturing, or more theHaltrim business there and
during the past.
We're up with a couple ofcompanies and trying to actually
generate more data from thevalue of the product.
Peter Warren (00:54):
Thanks very much and just bringing in some
information from our ongoingsurveys that we do with our
customers and other folks thatare not our customers in the
industry.
We call it our voice of theclient survey, so our voice of
the customer survey, I guess iscorrect.
We pointed out this year thatand this is not fully published
yet but there's a lot of peopleinterested in AI.
(01:16):
Obviously they're interested inautomation.
They're trying to do more withless.
They're trying to manage thefact that they're not being able
to get all the people they want.
They've got cutting budgets,they've got the fact that they
have people retiring withknowledge.
All of that is sort ofhappening all at the same time.
One of our customers said thiseverything's happening
(01:37):
everywhere all at once, and Ithink that's a quote from a
movie.
Perhaps you know, diane, what'syour thought on the most
exciting trends for energy andtransitions?
How do you see this AI actuallybeing able to fill those
problems I just described?
Diane Gutiw (01:51):
You know, I think we saw a lot in Hanover and I
know we all just were at theHanover Messe event and there is
so much going on.
It's hard to nail down onething Automation and the move to
agentic AI to extend what wewere already doing in
intelligent automation intoagentic AI is brilliant.
Also, the efficiencies thatorganizations are gaining
(02:15):
through use of some of thesegeneral purpose tools.
Peter Warren (02:17):
And Lucas.
I mean we also saw theconfluence between the logical
layer, which we just sort ofdescribed, but also the physical
layer of which we just sort ofdescribed, but also the physical
layer of devices like yourclients are using, but also the
overlap into physical security,just a ton of things, the use of
imagery, lidar, radar,overlapping and bringing really
(02:38):
a bigger holistic viewpoint topeople.
But what's your viewpoint onsort of what was the most
exciting thing that we picked upat Hanover and where you see
your customers going?
Lukas Krappman (02:46):
Yeah.
So, frank, what's interestingfrom, or what was interesting
from my perspective, was thatNow everything's basically
available in terms of technicalknowledge and the tools,
(03:20):
macroeconomical challenges,meaning how everything comes
together and how the wholeecosystem, for example, of
manufacturers, energy companies,logistic companies and also
health and life science, canreally work together to solve
the problems which are burningum on our plate right now.
Yeah, and this is, I think,especially due to the
geopolitical and other economicinfluences, really the thing
which is most interesting formost of our clients out there.
Peter Warren (03:42):
Yeah, you bring up a couple of good points.
There is that the supply chainsor the ecosystems are all under
attack.
People are having to re-vector.
What some people would take twoyears to do now has to happen
in like two days.
We've also heard from a voiceof the clients.
People are looking for new,better software tools to be more
agile in both their regulatoryand tariff type of management,
(04:04):
given what's going on.
So there's a lot of things thatpeople are trying to deal with.
Diane, how do you see AIplaying a role in all of that?
It seems very complex.
Diane Gutiw (04:14):
Yeah, it's a really good point, but when we come to
innovations, we're at a goodplace in having tools that are
allowing us to do things thatwere really complex, expensive,
complicated to do in the past.
You know, in addition to theeconomic climate globally, which
energy is absolutely at theforefront of, we also have a
rapidly aging populationglobally, and when Lucas is
(04:36):
talking about parallels acrossdifferent industries, that's a
challenge all industries aretrying to solve.
How can we do more, do morepersonalized, specialized type
work with less people being ableto do it?
And, as you mentioned earlieras well, a lot of the people
(04:57):
that have the deep knowledgeinto systems are retiring.
So, when we look at AI, some ofthese things that are the
challenge maintaining, retainingknowledge, upskilling people
quicker, being able to take oversome of the menial tasks so
people are working at the top oftheir skills rather than
spending days of their weeklooking for information,
collating information and tryingto do analysis.
(05:18):
There's tools that are notreplacing people, but are
replacing tasks, and replacingtasks that could be done much
quicker, perhaps moreefficiently, so that people are
able to do what they need to doto be creative.
I think across all industries,we're starting to see this
challenge and finding patternsand opportunities both to reduce
(05:39):
risk and improve efficiencies,and improve those opportunities.
Peter Warren (05:43):
Yeah, if I can ask an open-ended question to both
of you and you can let me knowwho would like to answer first,
the cross-industry thing.
We've hit it a couple times andwe're seeing at least from our
point of view as CGI, because welook across industry that the
problems we're trying to solvein one industry actually
parallel or have already beensolved in another industry.
And looking to the benefit ofthose things, whichever one of
(06:06):
you would like to go first,maybe you'd make a comment on
how we're actually learning fromother industries and pulling
things back in.
Diane Gutiw (06:12):
Sure, I'm happy to take that one because I think I
probably spend my time in themost different industries lately
.
I think there is bettercollaboration now.
Certainly you know ourselves atCGI.
We spend a lot of timesupporting the specialists
across the different industriesfor that very purpose.
It's a best practice in datascience as well is to not
(06:34):
reinvent the wheel, to findefficiencies and reuse patterns.
So a good example would bepredictive maintenance.
What we've done in drilling inthe past long predating
generative AI we've reused in anumber of different industries
to look at how to determinepotential faults in different
types of equipment when it'shard to nail it down.
(06:55):
So clustering that data andthen using a K-shape algorithm
to nail it down.
It doesn't matter if that'sjust mining or utilities or
anything that's using heavyassets or any assets you know,
even looking at modalities inhealthcare.
That type of model and approachto solving problems can be
shared.
So the population retiring isthe same thing.
(07:17):
Where are their efficienciesand where in a workflow, in a
value chain that we can enhancewhat we're doing with AI rather
than replace, so that the peoplecoming in are upskilled much
quicker.
People that are doing the workare able to focus on a
meaningful task rather thanmenial tasks.
That's a problem that we're alldealing with, and I think there
(07:37):
definitely are efficiencies.
Our vendor partners are good atsharing that across industries,
and I think that's one of thereal values that we have as our
team at CGI.
Peter Warren (07:46):
So, Lucas, at the Hanover Messe you ran a workshop
with customers on digital twinsand digital triplets.
What was sort of the keymessage that you took away from
that?
What were the clients lookingfor?
What were the clients lookingfor?
What were they interested inmost?
Lukas Krappman (07:59):
Yeah, I mean like it was a pretty interesting
workshop, to be honest.
So we've raised the workshoparound digital twins and
triplets in the hygieneecosystem, but the participants
were actually from a couple ofindustries.
So we had, for example,participants for the hydrogen
industry, we had manufacturers,but we also had participants,
(08:20):
for example, from agriculturalbackground, and the most
important thing all threebasically have in common is that
they kind of like have to makea lot of data available with
certain speed and and thengenerate insights from it.
So how do you connect, forexample, data coming from a
(08:42):
farmer's field, from crops, forexample, to an infrastructure
and add a business applicationon top so that you can do a
proper farming or, for example,put out water for the seeds if
there's rain in the weatherforecast, stuff like that?
This is, for example, for theagricultural perspective and
then for the energy andutilities industry.
(09:03):
For example, I had a couple ofcustomers with a more like a
hydrogen background and theirchallenge basically is on how
they can integrate the data from, for example, the hydrogen
electrolyzers then into, forexample, large energy grids, on
how to optimize productiondepending on the weather, do a
lot of predictive maintenancefor example, when do we have to
replace a particular stack.
(09:24):
Is it better, for example, tonowadays turn on my hydrogen
electrolyzer to generatehydrogen or the energy, or
should I look for like ananother source to do it?
Yeah, so it's actually aboutconnecting all the different
data points out there in thedifferent ecosystems and then
making an informed decisionbased upon it.
As you might know, there's alot of data out in the industry
(09:48):
and in the ecosystem.
There is a lot of metadata outof it.
That's even a more importantthing.
It's not only about thedifferent signals you have out
there, but also the metadata andhow you process it, and I think
this is going to be a realchallenge in the next couple of
years.
Peter Warren (10:06):
Lucas, I really appreciate what you said on that
because it really highlightsthe data and it was interesting
that a use case from what we'retalking about hydrogen was so
applicable to somebody in theagricultural industry.
And what they were trying to dowas an interesting dialogue
with the researchers and otherfolks that were in the session.
Just heading to wrap up hereand give you guys a chance to
sort of summarize what do youthink is the vision on the
(10:30):
global stage, what's the bigthing moving forward?
Where is this all going?
And maybe we'll go back to you,lucas, first and give you a
chance to give that summary andthen we'll toss it to Diane.
Lukas Krappman (10:40):
Yeah, I'm like when you were walking around
with all the different booths,right, so you saw that AI is
actually part of the thing.
Every booth.
I didn't see like one boothwhere AI wasn't incorporated or
incorporated into the businessmodel, right.
And this starts from, likeintelligent co-pilots, from
custom GPTs, for example,different use cases from energy
(11:02):
suppliers out there.
From my perspective, and whatwe talked about to different
customers.
It's really about how AI cancontribute to, for example, the
efficiency of each of ourcustomers out there, how their
day-to-day work actually can getmore efficient and bring more
benefit to them.
And with that, I guess I'llhead over to Diane.
Diane Gutiw (11:23):
Yeah, you know, I think the thing that's happened
in this last year is AI nowisn't just one thing, it's not
just generative AI that peoplehave on their phones.
I find that innovations aresplitting into three things, and
those three things are reallywhat we saw in Hanover, as Lucas
said, as well as what's goingto drive us forward.
The first is tools for internalefficiencies.
(11:45):
This is your operational,administrative access to
information, access to data tobe able to do things in your
day-to-day workflow quicker.
So a lot of what came out ofthe GPTs and software
development and being able tointegrate with everyday work
tools is going to be a hugeimpact on every sector, but will
(12:07):
be able to help a lot.
The second would be automation.
We already see a lot ofautomation in this sector, lots
of robotics and automation andprocess engineering being
advancing technologies in thisspace.
I think that by havingsolutions that are able to do
(12:28):
more, having more autonomousautomation, having automation
that's very focused on morecomplex tasks which were hard to
complete without really clearinstructions, you're able to do
more.
And then the last one is data.
We've talked about data a lot,but using AI to be able to mine
that data, to find patterns andhave more evidence-based answers
(12:50):
to questions.
You know, to me those threethings are splitting out of this
technology evolution that we'veseen and those are, I think,
what the three top areas thatare going to have a real
positive impact on the energysector.
Peter Warren (13:02):
Well, I'd like to thank you both, and I will add
one thought too is this was aconversation we had with other
clients.
Diane, when we were there, isusing generative AI to actually
error correct and check data andvalidate, so AI is actually
helping us improve our data inreal time.
So that's a great thing.
Well, thank you both, reallyappreciate this.
So hopefully everybody enjoyedthis session and we'll do more
(13:24):
coming up in the coming weeks.
Thank you, lucas.
Popular Podcasts
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.
24/7 News: The Latest
The latest news in 4 minutes updated every hour, every day.
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com