May 2, 2024 • 33 mins
Advancements in AI technology have made it possible to create virtual representations of our real-world environments, and these digital twins could change how we experience just about everything in our lives. These digital twins can create everything from more efficient workplaces to smarter traffic lights, and on the first episode of Season 2 of Technically Speaking: An Intel Podcast, digital twin expert Tony Franklin discusses what’s required to create a digital twin and how different industries can use this technology to create a safer future for everyone.
Learn more about how Intel is leading the charge in the AI Revolution at intel.com/AIeverywhere
See omnystudio.com/listener for privacy information.
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:04):
Welcome to tech Stuff, a production from iHeartRadio. Hey they're
tech Stuff fans. This is Jonathan Strickland speaking. I've got
something special for y'all today. What follows is an episode
from season two of our podcast technically Speaking. This particular
(00:24):
episode is about digital twins, which I thought was a
really fascinating topic. It was one that I wasn't actually
really familiar with, and I learned a lot and I
hope you will too. So check this out, and if
you enjoy it, make sure you go and subscribe to
technically Speaking. You've got two seasons of really cool material
(00:48):
in those episodes, so check that out. And I hope
you enjoy this episode.
Speaker 2 (00:55):
Where do world changing ideas get This start at Intel.
It starts with real solutions, and real solutions start with
exceptional engineering, the quantum computing revolution, the next generation of
AI experts, the renewable energy grid, liquid cooling, data centers,
early diagnosis for cancer, water restoration, and even farmland protection.
(01:15):
The examples are countless, the impacts are endless, but the
foundation is always the same. It starts with Intel. Join
us in redefining what's achievable through the power of AI.
Learn more at Intel Dot com slash stories. Workplaces like
factories or fulfillment centers are filled with many moving parts
(01:38):
and require constant supervision and alertness to ensure worker safety.
But what happens when errors or accidents happen unexpectedly On
the simplest level, it creates chaos and can impact productivity.
On a larger scale, can create a dangerous environment for employees.
How can technology like AI and the creation of a
(01:59):
digital twe in help quickly correct errors and prevent accidents
before they occur? And could these virtual replications of a
physical space ensure workers go home to their families safe
and sound. Join us as we learn more about the
world of digital twins and the many ways they can
not only improve workplace safety, but also public safety. Welcome
(02:25):
to Technically Speaking, an Intel podcast, the show that brings
you the stories and insights of AI, presented by iHeartMedia's
Ruby Studio and Intel. Hey then, I'm gram class. Today,
we're exploring the spaces and places replicated by digital twins.
For starters, what is a digital twin? We're going to
(02:46):
be examining digital spaces that represent an actual physical space
in our world. To discuss the topic further, We're joined
by Tony Franklin. Tony Franklin is the general manager of
the Federal and Aerospace Markets, which includes military, aerospace, and
government within the Network and Edge Group. He has more
(03:06):
than twenty years of corporate entrepreneurial, business development and management experience,
focusing on starting and growing multiple businesses that apply to
the Internet of Things, intelligence systems, and communications technologies. Intel's
Network and Edge Group provide solutions that lead the industry
and transforming businesses and the way we live. Are making
(03:27):
it simple to create exciting new diet solutions.
Speaker 3 (03:30):
Welcome to the show, Tony, Thank you, thank you, glad
to be here.
Speaker 2 (03:35):
We've seen a lot of science fiction depictions of digital
twins over the years and movies and films. The one
that comes to mind is the Matrix, where the entire
world is the digital twin. Although it's a useful sinister motives,
I'd like to get your definition of what a digital
twin is.
Speaker 3 (03:52):
Yeah. Sure, it's interesting you said the matrix.
Speaker 4 (03:54):
I had to laugh because of all the examples we've
joked about, that's one that hasn't come up, and it's
so obvious when you said, I'll start to use that
in its simplest form for me, a digital twin is
the digital replica of the real world.
Speaker 2 (04:07):
And in terms of the technologies that are out there
needed for digital twining, maybe you could describe a little
bit of how those sorts of systems are put together
and what are some of the technology that Intel has
that can help that.
Speaker 3 (04:25):
Yeah.
Speaker 4 (04:26):
Sure, it's really been an evolution of technologies that some
of them were all used to using, and some of them,
you know, if you're not in the field, maybe not
so much. And so Gartner, one of the well known analysts,
has this Emerging Technologies trending chart they do every year,
and they talk about edge AI, so artificial intelligence at
(04:46):
the edge, you know, the place where data is actually
being generated more so than in say the cloud or
data centers. That's happening, It continues to evolve, it's growing
more and more. We're pushing more and more intelligence to
the edge. And by the edge, it could be everything
from a self phone or refrigerator or a car.
Speaker 3 (05:01):
Again, where is the data actually being generated?
Speaker 4 (05:04):
And then they talk about digital twins being sort of
the now to three years. I think one of their
data points was something like forty percent of businesses large
businesses in particular plan to use digital twins over the
next two to three years to actually generate revenue, So
that's happening now, and then lastly over the next maybe
six plus years, they talk about the metaverse. Now, while
(05:25):
we don't generally talk about the metaverse as much, the
name means different things to different people, but it's the
full extent of how it was. Commerce and any other
business really leverage a fully digital space that has interaction
with the real world. So there's this spectrum that has
been happening between sensors, with cameras being the most obvious
(05:46):
because we're all using cameras today. Our phones have so
many censers we take them for granted, but all the sensors.
One of the key technologies that are needed an ability
to replicate if you're doing visual digital twins, so an
ability to replicate the real world, and of course physics
modeling if you're really doing analysis and you need to
replicate the physical asset in a digital world, So computing
(06:10):
capability to be able to replicate the behavior of the object,
with AI being a critical component to now actually apply
intelligence and analysis to the twin. So when you think
about those different areas well. Intel, we don't make sensors.
Everything else along the way is where we tend to
play clearly, computing technology, the ability to apply AI, so
(06:33):
both from the software side and the various different computing
technologies that enable AI, whether it's processors or GPUs, are
AI accelerators, et cetera. And then of course we have
a very broad, really world class partnership and ecosystem that
we work with to enable the different industries.
Speaker 2 (06:54):
Okay, and in terms of trying to get like a
visual kind of representation of what this would clients say, say,
if you're walking in a warehouse, for example, and you're
looking at the shelves around you, you might see some
conveyor belts. How would it actually look in a digital twin.
It depends on the use case. The simplest version i'd
(07:15):
use that many people should be able to relate to Obviously,
many people can relate to matrix. But from an application standpoint,
I would say think about Google Earth or Google Maps.
Even that is a type of model right. Another example
is many of the retail applications allow you to basically
(07:35):
embed their items that are for sale into the digital
replication of your particular space. So it's a real world
and digital combination. That's always the key.
Speaker 4 (07:47):
So those are very basic, simple applications that people use
and don't even realize. They don't think about them as
digital twins, but they're already getting used to this relationship
between the real world and the digital world.
Speaker 3 (07:59):
Yes, I saw.
Speaker 2 (08:00):
Intel has a software platform called Scenescape that can transform
data from this world of senses to create a real
time digital twin of your physical space. Can you tell
me a little bit more of how that works? So
there's really three basic steps. One is mapping your space.
The key though in the type of real time digital
twinning that we pursue is it is a coordinate, accurate space,
(08:23):
so it's not just taking a random space. We know
that that room is twenty feet by twenty five feet,
and we also know that the cameras are six feet
up on the wall. We know the actual XYZ of
the space, So that's the first step. Second step is
calibrate the space. So calibrate meaning I have the space,
where are my sensors? So my sensor is ten feet up,
(08:45):
four feet over in that corner, et cetera. Now you
turn on your sensors and you ingest that into scenescape.
Notice I haven't actually visualized necessarily anything. So the key
for people to realize in real time digital twinning. There
are some applications where someone's not actually going to be
sitting there watching the digital twin of the store. They
don't need to do that real time. The AI and
(09:09):
computing obviously, the AI tools and the actual AI models,
the inferencing capability of scenescape is doing the work. You
are configuring it so you could add things like heat
maps or trip wise so that you can actually have
events based on whatever policy you want to implement, so
(09:30):
that you don't have to actually monitor. So I'll know
that over in this area of the meat department, if
there's more than twenty people for twenty seconds, there's something
that happens. I don't even have to watch anything for that.
The intelligence is making it happen now. If what I
want to do later is now hit the rewind button.
(09:51):
The founder and creator for this particular product, he has
a phrase I'd love to use. It's called the DVR
for the real world. AI is happening real time, but
if you want additional analysis afterwards, you have that capability.
I can imagine there's a lot of challenges trying to
come up with these digital twins. What are some of
the top challenges or issues that people who are looking
(10:14):
to try and deploy these sorts of systems would have
to consider.
Speaker 4 (10:19):
The most prominent one, to be honest with you, is
the mindset more than anything technical. You think about some
of the technology that's growing in our own home, siy
alexa are cars. There's so much technology, and most people
really don't understand you're already using AI. In many cases,
(10:40):
you're already using some sort of digital twin technology. There
was one demo we had for Scenescape and the executive
loved it. It's like, this is amazing. I can do
motion tracking. I see where people are. I can have
multiple cameras monitoring the same asset, our person or object,
but I only see one, so it deduplicates the person.
I can track with somebody's been in a space. Maybe
(11:01):
I have a radiation sensor and I can actually track
how long that person has been in the space, and
I can set triggers. There's so much he saw that
can be done, and he was so excited and he said,
where's the AI, right, Well, it's the AI that's doing
everything you just described. That's right, you just and it
actually set us back for a second. We're like, well, clearly,
(11:22):
we need to make sure we understand where people are
starting from. We can't assume they already know there's a
level of technology and integration of their technology.
Speaker 2 (11:33):
And that's one of the biggest challenges when it comes
to understanding digital twin technology. It's the messaging some of
the very tools that you're accustomed to right now, like
the cell phone or the smart speaker that you are
listening to this podcast with essential when we consider the
future of digital twining. So when it comes to a
future that incorporates AI into our daily lives, we've actually
(11:54):
already taken the first steps down that path. One of
the things I like to examine is the way that
technology actually helps democratize. And maybe you have some sense
of the type of customers. Are they sort of large enterprises,
because I'm really keen to see this sort of technology
(12:15):
really get pushed down to the smaller businesses, you know,
and make it affordable for them to adopt and use.
Do you have any thoughts about that particular trend.
Speaker 4 (12:24):
Yeah, absolutely, I'd say they're all generally larger enterprises, but
they may be larger enterprises with smaller facilities, so they
have to think about the implementation at the store level,
and then they can step back and look at it
at an operational level for the entire business that they're
trying to run. When you think about physical security, well,
(12:45):
physical security can happen on a construction site, it can
happen in an office space, it can happen anywhere. But
the companies we're dealing with are generally the companies that
one have the actual technology, so they may be the
camera vendors, et cetera. But whereas actually being implemented the
targeting a broad range in particular segments like I mentioned,
but the actual implementation may happen at a different level.
(13:07):
So it's the companies that apply technology across specific segments
and then they actually tear those down. Cities can be
large or they can be smaller, but you're implementing generally
starting at an intersection level, so that could be maybe
four cameras max. But now I've got a thousand intersections,
so it grows in scales, and what we're seeing back
(13:30):
to that early adopter, we see the big picture.
Speaker 3 (13:32):
Let's start with.
Speaker 4 (13:33):
Three intersections, and let's see and understand where technology can
be applied there. Because one of the ways I like
to explain to people, you need to understand the environment
the scene. That's what it's called scenes. Gape the scene,
the area, your environment better. That's one way to think
about digital twin being able to enable that.
Speaker 2 (13:54):
Coming out next on Technically Speaking and Intel podcast, I.
Speaker 4 (13:58):
Am the ultimate digital twin that I watch. I don't
care about an avatar that's fun and fancy. I was
something that helps improuse my quality of light.
Speaker 2 (14:07):
We'll be right back after a brief message from our
partners at Intel. Where do world changing ideas get their start?
At Intel? It starts with real solutions, and real solutions
start with exceptional engineering. Empowering those with disabilities starts with
(14:30):
assistive AI, and stopping crop loss from infestation starts with
thermal imaging and open technology, while artificial intelligence that predicts
depression starts with educational programs like Intel's AI for Youth.
And that's just the start the quantum computing revolution. The
next generation of AI experts the renewable energy grid, liquid cooling,
(14:54):
data centers, radiation exposure prevention in space, water restoration, and
early cancer detection. The examples are countless, the impacts are endless,
but the foundation is always the same. It starts with Intel.
Learn more at Intel dot com, Forward Slash Stories. Welcome
(15:20):
back to Technically Speaking, an Intel podcast. I'm here now
with Intel's own Tony Franklin. Do you have any other
examples of benefits that your customers have seen, whether it
be productivity, increase, revenue, better, safety.
Speaker 4 (15:37):
Yeah, I'll go on reverse since you said safety last,
because that's one that is so common yep to people.
In fact, I think it was the university in Texas
that has doing some pilots with the cities and with
smart vehicles. And it's a device called a roadside unit. Again,
most people don't even realize that you pull up to
an intersection, there's normally a smaller box on the side.
(15:59):
You already have the box that controls the lights, etc. Well,
I want to do more so you can make that
unit more intelligent. You can actually allow that roadside unit
to communicate with cars. As cars become more intelligent, they
have five G, they have wireless communications. So they implemented
a pilot where there was a particular intersection, so as
the car pulls up. Imagine an alley off to the left,
(16:20):
so the car can't see down the alley clearly, but
there's a pole on the right that has a camera.
The camera can see the car coming. The camera can
see down the alley. The camera has a roadside unit
with intel processing equipment. It's running scene skate again. They
don't even need to visualize this. The camera sees someone
walking down the alley, the car is coming forward. It
(16:43):
can communicate to the car because even the cars with
the cameras can't see around corners, so the camera can
communicate there is somebody walking. You need to slow down.
Knowing the speed of the car is great acceleration, we
understand that, but knowing where that car is at the
speed of a car coming down the highway is one thing.
The speed of a car coming down that street where
there's an alley is a totally different scenario. I need
(17:05):
to know the location of that car relative to the
camera and relative to the person around the corner, both
coming at the same time. So three D is also
a key aspect and value of digital twin That translates
to end benefit like you're talking about. So that's safety
right there, and that safety translates to insurance as an example.
Speaker 2 (17:23):
Yeah, just fitting off that safety theme. Can technologies like
scenescape and having those sort of cameras help with worker safety,
say in a factory or warehouse, where they can detect
or even predict, you know, if something's going to go
wrong and actually warn a worker that something's going to happen.
Speaker 3 (17:46):
Yeah. Absolutely.
Speaker 4 (17:47):
Robot interaction is a common one also, so think about
robots with cameras and the cameras and sensors that are
around Mobile World Congress is going on right now, and
I think it was last year. We did a escape
demo there and it was purely an industrial We had
the robotic arms that were moving and they were building something.
And then you have a sensor doesn't even need to
(18:09):
be a camera that has a digital n scen escape
the trip wire, so we know if somebody crosses this point,
then it's a trip wire. So that was the actual demo.
So there was a safety zone and then there was
crossing the safety zone. So if you enter the safety zone,
there could be a warning light to go off. You
don't have to stop anything, but I know someone's in
the safety zone, I know how long they've been in
the safety zone and if they cross past that, then
(18:30):
I know I can start to shut down automatically equipment
if that's the policy that that particular site chooses to use.
So I think can execute another one in a more
constrained environment warehouse that we've seen is where there's actually
a controlled space, so radiation. Actually, the earlier example I
talked about is a real example where there's an area
(18:53):
that it needs to be climate controlled and it literally
has radiations. So they have a radiation sensor both inside
outside and commodity or do you have the equipment on
how long has a person been in this particular space,
and I could set timers and triggers so I know
that they can only be in for so long, and
I can also track that. So that's real time action
and control, and I can also use that for later
(19:15):
analysis and prediction. Maybe I need to change the configuration
of the room, maybe I need to put more signs up,
But you can have real time action and decisions and
also post analysis.
Speaker 2 (19:25):
Yeah, what you said there about the simulation is quite
interesting because you know, as you're talking else, you know
it came back to the gaming side of things, playing
SimCity or roller coaster tycoon being able to sort of
simulate you know, if I put this thing here, is
it going to be dangerous?
Speaker 3 (19:42):
If I put that over there? Does that help the workers?
Does it help with productivity?
Speaker 2 (19:46):
Maybe talk a little bit of some examples of using
real world data to kind of do what if analysis
of various scenarios that management and workers together can can
see mille and potentially improve the workplace.
Speaker 4 (20:03):
So let's take something like a gaming site, I mean
like a football or soccer So clearly those are massive
events with a lot of people, a lot of insurances,
there's safety concerns, there's access to medical professionals that need
to get in and out. So can I take existing
data that has already been captured using existing cameras and
(20:27):
I can actually run simulations on that. I could also
ideally what I want and I need it. If I'm
going to do a digital twin, I need some sort
of digital twin of the environment. The level of depth
is just depending upon the level of analysis that you
want to conduct. Now, what I need is what's the
data that I've been able to collect, because most of
these places they're already going to have some data, even
if it's just camera feed data. I could take that
(20:49):
and actually start to run models on Okay, where are
people congregating? I can actually post camera feed and apply
inference data to that, so I can use the AI
to it. Andy, well, that's a person, and that's an animal.
That's a car over there, And now I can start
to look at Okay, how often are they in these spaces?
Where am I getting congregation? Where am I getting long
(21:10):
killing live? So I can do analysis all on existing data.
Now I can start to reconfigure whatever actions need to
be taken, so all of that can happen before I've
shown up physically at the space.
Speaker 2 (21:24):
Just think about all the personal identifiable information involved in
some of the tasks we're talking about today. Well, the
sheer amount of streaming data coming in from a host
of senses required to implement digital twining, keeping that data
secure is paramount to the future of this industry. I
like to get your thoughts around the whole privacy side
of things, and you know what can be done to
(21:46):
make sure that as individuals we don't feel like where
our privacy is getting invaded.
Speaker 4 (21:52):
That's a very good topic and we thought about that
from the beginning. So one of the ways that we've
defined scenescaper is we primarily work on meta data, and
what metadata simply means is, for instance, we don't do
any facial recognition. I need to know that that's a
person or I need to know In fact, we had
an actual scenario where a customer had a particular area
and they knew people were around, but at night there
(22:13):
were objects and they didn't know what it was. They
were animals, and the model hadn't been trained for animals.
So the model can say, hey, there's something there. I
can't say that it's a deer versus or whatever, but
it's not a human, you know. And so think about
the simplicity of that. Now, I don't have to transmit
every movement, every aspect. I'm only transmitting what's critical to
(22:35):
make the decisions that are needed real time and for
post analysis.
Speaker 2 (22:40):
We talked a little bit about fulfillment centers and warehouses.
I like everyone else's ecomma sites like Amazon Prime. I'm
just wondering if you could maybe paint a picture of
how from the time that I hit that buy now,
but to the time that I get my pair of
socks at my doorstep, take me through how digital twins
(23:02):
could be used. How would a system help that process,
both as an in consumer and also for the business.
Speaker 4 (23:11):
Actually to be honest, One of the first examples that
came to mind is the delivery truck and why location
intelligence is so important. All of us use location based
services today. There was a study I read I think
it was UPS is saving a lot of money per
truck because they realized that location intelligence they were getting more.
(23:33):
Particularly this was when hotels were putting in their address
because they use Amazon Prime to and they're getting these
packages shipped to them. The location data of that hotel
relative to where the truck is coming from, and then
mapping the route were not good routes, so it was
costing the company so much money to get from point
(23:55):
A to point B. So now I can start to
identify where are the hotel. And they discovered this and
they started taking copies which they have.
Speaker 3 (24:02):
This would go.
Speaker 4 (24:03):
We can take copies of the maps, I can start
to locate where am I going. I can start to
figure out the routes. So they're using the twin of
the maps and the data they already have. Think about
they have tons of data on their routes and the
locations and where are they normally congregating, and which truck
should they send, what time should they be They did
all of that analysis to figure out just on the
(24:25):
back end of when I actually dropped the packets off
to you and what makes sense that's saving money for
them completely and again it's location based yep.
Speaker 2 (24:34):
And can you give me an example of how digital
twinning might already be in use for the consumer on
one of these sites such as Amazon Prime all similar.
Speaker 4 (24:43):
I have a chair behind me I just bought, so
now I can use digital twining right now to figure
out exactly where I want this, how does it look?
And they also have those clothing services which are digital twining.
You're the real person, and they have the digital where
you can apply the clothes to you. Yes, I mean again,
these are services that people are using today. But back
to my earlier comment, you're not actually thinking about the technology.
(25:07):
Are taking that now back to work, to your day job. Oh,
I do all of this at home. I should be
applying this to my business and saving money and getting
greater insights of my scene and of my environment. So
those are a couple of examples. Can I give you
a different example. Most people have some sort of ring
doorbell or type of it could be ring, it could
be simply safe whatever.
Speaker 2 (25:28):
Yeah, I'll have that.
Speaker 3 (25:29):
There you go.
Speaker 4 (25:31):
One of the ways we've gone to market is to
make sure what we're doing is standard based and open,
you know, maximum scalability and flexibility. So I have a
ring one of my family members has simply safe. The
challenge is I have the ring camera at the door,
I have another ring camera. I can connect them. I
can see if something's walking by. That's great. It's not
(25:53):
very open. I'm somewhat siloed. What we've had someone do
with scenescape is they use scenescape. First of all, you
don't need to go through the so they're not paying anybody. Okay,
if you want to use the cloud, you can, but
you do not have to use the cloud. Everything can
be edge based, and by edge based, think about the
edge again. At home, where's the data generated? That's my edge.
So in this case, your edge is your home. So
(26:13):
my home, I already have a computer, and I already
have one or two cameras. But there's a particular type
of camera I want for the front yard, which is
totally different than the camera I want indoors, which is
totally different than the camera I want, so three totally
different brands. Well it's called multi camera multi brand from
that sense. So and by the way, maybe I want
a heat sensor or something like in a particularly air
in the backyard because I don't know if there's something overheating.
(26:35):
So now I can add all different type of brand
sensors and I can connect that into scenescape. And now
scenescape party has AI. So I've used different brands. I've
used my own computer, so I have standard based connectivity.
And because of standard based connectivity, I can connect it
to my phone. Every phone app now it's very easy
to connect to it and get alerts. So now I
(26:55):
can start to use the existing AI tools that are
in Scenescape. But there are so many applications out there.
With scenescape, you can integrate it with other applications. So
there was one person that used it to identify the
difference between a car coming in the driveway versus a
postal truck that goes by and stops, and they set
an alert. So whenever the postal truck comes and stops
(27:18):
for a few minutes, the alert goes on the phone.
He never has to look at a camera. He knows
when he gets that alert mails here. You can't do
that with RING, you can't do that with these other applications.
So that's a common use case that people know today
where standard digital twinting technology with AI, standard based communication,
and standard computing technologies can all be used to enable
(27:42):
use cases that we use every day.
Speaker 2 (27:44):
Yep, we just have time for one more question. I
would like to get your number one. I guess area
of excitement for digital twins for.
Speaker 4 (27:57):
Me is healthcare. What I want to see in my lifetime,
and we have the technology to do it. In fact,
we've have a few use cases with scene skates where
we're working with the medical community. I want the digital
twin of my health. I want for me the person
all think about all the data, all the medical records.
Speaker 3 (28:14):
First of all, it's hard enough keeping all your medical
records together.
Speaker 4 (28:16):
So not only my medical records, but the medications I've taken,
any reactions I've had. You have so much data from
blood work and positive reactions, negative reactions to medications, exercises
I've done that may have improved weight or blood pressure.
So as I grow and as all of us grow
and age, I should say yes, I want all of
(28:40):
that history to follow my DNA, my person to maximize healthcare.
I am the ultimate digital twin that I want. I
don't care about an avatar that's fun and fancy. I
want something that helps to improve my quality of life.
That's what I want out of interest. Have you seen
any companies or businesses looking into this. I did meet
(29:00):
a company or CEO of a company that's working on
the medical record side where they're trying to tie all
of that to the person so that can follow them,
so that now physicians and healthcare workers can have all that.
So it's a startup. It seems to be much more
challenging to get this done than you think it would be.
But we've engaged with some companies and hospitals that are
(29:22):
making their hospital smart. You can see some areas called
the smart operating room. That's a particular area in a
hospital that's obviously critical. I mean, you think about something
as basic as we're in the operating room, we're starting
the operation. I have twelve high value instruments on my right.
Those twelve high value instruments need to be there when
(29:42):
I finished, because if they're not there, there's a very
bad place they could be. Ye, that's right, and that
is a real example that I know personally somebody like
that that's happened to and they've had to go back
and get one of those instruments. So when you think
about the seriousness of the operating room, and that's before
you even get into intrusive sensors and I mean, you know,
what's the blood pressure and et cetera. Yes, you never
(30:04):
go to a hospital, are to a healthcare professional and
they take your blood pressure and that's it and you're good.
Speaker 3 (30:09):
Then they start talking to you.
Speaker 4 (30:10):
No, we don't even think about the fact that we
take blood pressure, we take temperature, we take weight, sometimes
we take blood. So we're already experiencing a multi modal
environment to maximize our health, but we don't always think
about that when we bring that to work. So now
I need a temperature sensor, I need light, I need
lie dar, I need cameras, I need different brands. I
need to apply intelligence to that. So now I can
(30:32):
perceive my space and my environment, I can understand it
with analysis and AI and make sense of it to
make decisions, and then I can also do prediction based
on that. So what should happen in the future.
Speaker 2 (30:43):
That's great, Tony. I think we will leave it on
that note. Thanks so much, No.
Speaker 3 (30:48):
Thank you, this was fun, really enjoyed it.
Speaker 2 (30:52):
My deepest thanks to Tony Franklin for sharing his equities
with us. Today's chat about digital twins really our eyes
to the incredible potential. It's like stepping into a simulation
game where you can tweak maintenance schedules, production lines, and
even play around with the interaction between workers and machinery
using real world data. Yes, I'm letting my inner geek
(31:14):
shine through here, but the idea of managing a supply
chain with the ears of playing SimCity seems pretty cool
to me. Tony's closing thoughts on the future of healthcare
and the possibility of creating a human digital twin we're
particularly striking. Imagine having a clone of yourself in a sense.
I mean, we're already wearing watches that monitor heart rate,
(31:35):
physical activity, sleep quality, plus a range of other biometric data.
It's not too far fetched to dream about a future
where digital twins can forecast our health outcomes based on
our DNA, diet, and exercise. It's an interesting idea and
I'm looking forward to seeing where this technology takes us,
(31:55):
and lucky for us, we'll get a chance to explore
this further on our next episode Tuesday, April twenty third,
on technically Speaking, an Intel podcast, we'll be learning about
some of the revolutionary implementations of AI in the healthcare
space with team members from Intel and Siemens Healthineers See
you then. Technically Speaking was produced by Ruby Studio from
(32:23):
iHeartRadio in partnership with Intel and hosted by me Graham Class.
Our executive producer is Molly Sosher, Our EP of Post
Production is James Foster, and our supervising producer is Nikia Swinton.
This episode was edited by Sierra Spreen and was written
by Molly Sosher and Nick Firshall. Where do world changing
(32:51):
ideas get their start? At Intel? It starts with real solutions,
and real solutions start with exceptional engineering, the quantum computing,
the next generation of AI experts, the renewable energy grid,
liquid cooling, data centers, early diagnosis for cancer, water restoration,
and even farmland protection. The examples are countless, the impacts
(33:12):
are endless, but the foundation is always the same. It
starts with Intel join us in redefining what's achievable through
the power of AI. Learn more at intel dot com.
Slash stories
Welcome to tech Stuff, a production from iHeartRadio. Hey they're
tech Stuff fans. This is Jonathan Strickland speaking. I've got
something special for y'all today. What follows is an episode
from season two of our podcast technically Speaking. This particular
(00:24):
episode is about digital twins, which I thought was a
really fascinating topic. It was one that I wasn't actually
really familiar with, and I learned a lot and I
hope you will too. So check this out, and if
you enjoy it, make sure you go and subscribe to
technically Speaking. You've got two seasons of really cool material
(00:48):
in those episodes, so check that out. And I hope
you enjoy this episode.
Speaker 2 (00:55):
Where do world changing ideas get This start at Intel.
It starts with real solutions, and real solutions start with
exceptional engineering, the quantum computing revolution, the next generation of
AI experts, the renewable energy grid, liquid cooling, data centers,
early diagnosis for cancer, water restoration, and even farmland protection.
(01:15):
The examples are countless, the impacts are endless, but the
foundation is always the same. It starts with Intel. Join
us in redefining what's achievable through the power of AI.
Learn more at Intel Dot com slash stories. Workplaces like
factories or fulfillment centers are filled with many moving parts
(01:38):
and require constant supervision and alertness to ensure worker safety.
But what happens when errors or accidents happen unexpectedly On
the simplest level, it creates chaos and can impact productivity.
On a larger scale, can create a dangerous environment for employees.
How can technology like AI and the creation of a
(01:59):
digital twe in help quickly correct errors and prevent accidents
before they occur? And could these virtual replications of a
physical space ensure workers go home to their families safe
and sound. Join us as we learn more about the
world of digital twins and the many ways they can
not only improve workplace safety, but also public safety. Welcome
(02:25):
to Technically Speaking, an Intel podcast, the show that brings
you the stories and insights of AI, presented by iHeartMedia's
Ruby Studio and Intel. Hey then, I'm gram class. Today,
we're exploring the spaces and places replicated by digital twins.
For starters, what is a digital twin? We're going to
(02:46):
be examining digital spaces that represent an actual physical space
in our world. To discuss the topic further, We're joined
by Tony Franklin. Tony Franklin is the general manager of
the Federal and Aerospace Markets, which includes military, aerospace, and
government within the Network and Edge Group. He has more
(03:06):
than twenty years of corporate entrepreneurial, business development and management experience,
focusing on starting and growing multiple businesses that apply to
the Internet of Things, intelligence systems, and communications technologies. Intel's
Network and Edge Group provide solutions that lead the industry
and transforming businesses and the way we live. Are making
(03:27):
it simple to create exciting new diet solutions.
Speaker 3 (03:30):
Welcome to the show, Tony, Thank you, thank you, glad
to be here.
Speaker 2 (03:35):
We've seen a lot of science fiction depictions of digital
twins over the years and movies and films. The one
that comes to mind is the Matrix, where the entire
world is the digital twin. Although it's a useful sinister motives,
I'd like to get your definition of what a digital
twin is.
Speaker 3 (03:52):
Yeah. Sure, it's interesting you said the matrix.
Speaker 4 (03:54):
I had to laugh because of all the examples we've
joked about, that's one that hasn't come up, and it's
so obvious when you said, I'll start to use that
in its simplest form for me, a digital twin is
the digital replica of the real world.
Speaker 2 (04:07):
And in terms of the technologies that are out there
needed for digital twining, maybe you could describe a little
bit of how those sorts of systems are put together
and what are some of the technology that Intel has
that can help that.
Speaker 3 (04:25):
Yeah.
Speaker 4 (04:26):
Sure, it's really been an evolution of technologies that some
of them were all used to using, and some of them,
you know, if you're not in the field, maybe not
so much. And so Gartner, one of the well known analysts,
has this Emerging Technologies trending chart they do every year,
and they talk about edge AI, so artificial intelligence at
(04:46):
the edge, you know, the place where data is actually
being generated more so than in say the cloud or
data centers. That's happening, It continues to evolve, it's growing
more and more. We're pushing more and more intelligence to
the edge. And by the edge, it could be everything
from a self phone or refrigerator or a car.
Speaker 3 (05:01):
Again, where is the data actually being generated?
Speaker 4 (05:04):
And then they talk about digital twins being sort of
the now to three years. I think one of their
data points was something like forty percent of businesses large
businesses in particular plan to use digital twins over the
next two to three years to actually generate revenue, So
that's happening now, and then lastly over the next maybe
six plus years, they talk about the metaverse. Now, while
(05:25):
we don't generally talk about the metaverse as much, the
name means different things to different people, but it's the
full extent of how it was. Commerce and any other
business really leverage a fully digital space that has interaction
with the real world. So there's this spectrum that has
been happening between sensors, with cameras being the most obvious
(05:46):
because we're all using cameras today. Our phones have so
many censers we take them for granted, but all the sensors.
One of the key technologies that are needed an ability
to replicate if you're doing visual digital twins, so an
ability to replicate the real world, and of course physics
modeling if you're really doing analysis and you need to
replicate the physical asset in a digital world, So computing
(06:10):
capability to be able to replicate the behavior of the object,
with AI being a critical component to now actually apply
intelligence and analysis to the twin. So when you think
about those different areas well. Intel, we don't make sensors.
Everything else along the way is where we tend to
play clearly, computing technology, the ability to apply AI, so
(06:33):
both from the software side and the various different computing
technologies that enable AI, whether it's processors or GPUs, are
AI accelerators, et cetera. And then of course we have
a very broad, really world class partnership and ecosystem that
we work with to enable the different industries.
Speaker 2 (06:54):
Okay, and in terms of trying to get like a
visual kind of representation of what this would clients say, say,
if you're walking in a warehouse, for example, and you're
looking at the shelves around you, you might see some
conveyor belts. How would it actually look in a digital twin.
It depends on the use case. The simplest version i'd
(07:15):
use that many people should be able to relate to Obviously,
many people can relate to matrix. But from an application standpoint,
I would say think about Google Earth or Google Maps.
Even that is a type of model right. Another example
is many of the retail applications allow you to basically
(07:35):
embed their items that are for sale into the digital
replication of your particular space. So it's a real world
and digital combination. That's always the key.
Speaker 4 (07:47):
So those are very basic, simple applications that people use
and don't even realize. They don't think about them as
digital twins, but they're already getting used to this relationship
between the real world and the digital world.
Speaker 3 (07:59):
Yes, I saw.
Speaker 2 (08:00):
Intel has a software platform called Scenescape that can transform
data from this world of senses to create a real
time digital twin of your physical space. Can you tell
me a little bit more of how that works? So
there's really three basic steps. One is mapping your space.
The key though in the type of real time digital
twinning that we pursue is it is a coordinate, accurate space,
(08:23):
so it's not just taking a random space. We know
that that room is twenty feet by twenty five feet,
and we also know that the cameras are six feet
up on the wall. We know the actual XYZ of
the space, So that's the first step. Second step is
calibrate the space. So calibrate meaning I have the space,
where are my sensors? So my sensor is ten feet up,
(08:45):
four feet over in that corner, et cetera. Now you
turn on your sensors and you ingest that into scenescape.
Notice I haven't actually visualized necessarily anything. So the key
for people to realize in real time digital twinning. There
are some applications where someone's not actually going to be
sitting there watching the digital twin of the store. They
don't need to do that real time. The AI and
(09:09):
computing obviously, the AI tools and the actual AI models,
the inferencing capability of scenescape is doing the work. You
are configuring it so you could add things like heat
maps or trip wise so that you can actually have
events based on whatever policy you want to implement, so
(09:30):
that you don't have to actually monitor. So I'll know
that over in this area of the meat department, if
there's more than twenty people for twenty seconds, there's something
that happens. I don't even have to watch anything for that.
The intelligence is making it happen now. If what I
want to do later is now hit the rewind button.
(09:51):
The founder and creator for this particular product, he has
a phrase I'd love to use. It's called the DVR
for the real world. AI is happening real time, but
if you want additional analysis afterwards, you have that capability.
I can imagine there's a lot of challenges trying to
come up with these digital twins. What are some of
the top challenges or issues that people who are looking
(10:14):
to try and deploy these sorts of systems would have
to consider.
Speaker 4 (10:19):
The most prominent one, to be honest with you, is
the mindset more than anything technical. You think about some
of the technology that's growing in our own home, siy
alexa are cars. There's so much technology, and most people
really don't understand you're already using AI. In many cases,
(10:40):
you're already using some sort of digital twin technology. There
was one demo we had for Scenescape and the executive
loved it. It's like, this is amazing. I can do
motion tracking. I see where people are. I can have
multiple cameras monitoring the same asset, our person or object,
but I only see one, so it deduplicates the person.
I can track with somebody's been in a space. Maybe
(11:01):
I have a radiation sensor and I can actually track
how long that person has been in the space, and
I can set triggers. There's so much he saw that
can be done, and he was so excited and he said,
where's the AI, right, Well, it's the AI that's doing
everything you just described. That's right, you just and it
actually set us back for a second. We're like, well, clearly,
(11:22):
we need to make sure we understand where people are
starting from. We can't assume they already know there's a
level of technology and integration of their technology.
Speaker 2 (11:33):
And that's one of the biggest challenges when it comes
to understanding digital twin technology. It's the messaging some of
the very tools that you're accustomed to right now, like
the cell phone or the smart speaker that you are
listening to this podcast with essential when we consider the
future of digital twining. So when it comes to a
future that incorporates AI into our daily lives, we've actually
(11:54):
already taken the first steps down that path. One of
the things I like to examine is the way that
technology actually helps democratize. And maybe you have some sense
of the type of customers. Are they sort of large enterprises,
because I'm really keen to see this sort of technology
(12:15):
really get pushed down to the smaller businesses, you know,
and make it affordable for them to adopt and use.
Do you have any thoughts about that particular trend.
Speaker 4 (12:24):
Yeah, absolutely, I'd say they're all generally larger enterprises, but
they may be larger enterprises with smaller facilities, so they
have to think about the implementation at the store level,
and then they can step back and look at it
at an operational level for the entire business that they're
trying to run. When you think about physical security, well,
(12:45):
physical security can happen on a construction site, it can
happen in an office space, it can happen anywhere. But
the companies we're dealing with are generally the companies that
one have the actual technology, so they may be the
camera vendors, et cetera. But whereas actually being implemented the
targeting a broad range in particular segments like I mentioned,
but the actual implementation may happen at a different level.
(13:07):
So it's the companies that apply technology across specific segments
and then they actually tear those down. Cities can be
large or they can be smaller, but you're implementing generally
starting at an intersection level, so that could be maybe
four cameras max. But now I've got a thousand intersections,
so it grows in scales, and what we're seeing back
(13:30):
to that early adopter, we see the big picture.
Speaker 3 (13:32):
Let's start with.
Speaker 4 (13:33):
Three intersections, and let's see and understand where technology can
be applied there. Because one of the ways I like
to explain to people, you need to understand the environment
the scene. That's what it's called scenes. Gape the scene,
the area, your environment better. That's one way to think
about digital twin being able to enable that.
Speaker 2 (13:54):
Coming out next on Technically Speaking and Intel podcast, I.
Speaker 4 (13:58):
Am the ultimate digital twin that I watch. I don't
care about an avatar that's fun and fancy. I was
something that helps improuse my quality of light.
Speaker 2 (14:07):
We'll be right back after a brief message from our
partners at Intel. Where do world changing ideas get their start?
At Intel? It starts with real solutions, and real solutions
start with exceptional engineering. Empowering those with disabilities starts with
(14:30):
assistive AI, and stopping crop loss from infestation starts with
thermal imaging and open technology, while artificial intelligence that predicts
depression starts with educational programs like Intel's AI for Youth.
And that's just the start the quantum computing revolution. The
next generation of AI experts the renewable energy grid, liquid cooling,
(14:54):
data centers, radiation exposure prevention in space, water restoration, and
early cancer detection. The examples are countless, the impacts are endless,
but the foundation is always the same. It starts with Intel.
Learn more at Intel dot com, Forward Slash Stories. Welcome
(15:20):
back to Technically Speaking, an Intel podcast. I'm here now
with Intel's own Tony Franklin. Do you have any other
examples of benefits that your customers have seen, whether it
be productivity, increase, revenue, better, safety.
Speaker 4 (15:37):
Yeah, I'll go on reverse since you said safety last,
because that's one that is so common yep to people.
In fact, I think it was the university in Texas
that has doing some pilots with the cities and with
smart vehicles. And it's a device called a roadside unit. Again,
most people don't even realize that you pull up to
an intersection, there's normally a smaller box on the side.
(15:59):
You already have the box that controls the lights, etc. Well,
I want to do more so you can make that
unit more intelligent. You can actually allow that roadside unit
to communicate with cars. As cars become more intelligent, they
have five G, they have wireless communications. So they implemented
a pilot where there was a particular intersection, so as
the car pulls up. Imagine an alley off to the left,
(16:20):
so the car can't see down the alley clearly, but
there's a pole on the right that has a camera.
The camera can see the car coming. The camera can
see down the alley. The camera has a roadside unit
with intel processing equipment. It's running scene skate again. They
don't even need to visualize this. The camera sees someone
walking down the alley, the car is coming forward. It
(16:43):
can communicate to the car because even the cars with
the cameras can't see around corners, so the camera can
communicate there is somebody walking. You need to slow down.
Knowing the speed of the car is great acceleration, we
understand that, but knowing where that car is at the
speed of a car coming down the highway is one thing.
The speed of a car coming down that street where
there's an alley is a totally different scenario. I need
(17:05):
to know the location of that car relative to the
camera and relative to the person around the corner, both
coming at the same time. So three D is also
a key aspect and value of digital twin That translates
to end benefit like you're talking about. So that's safety
right there, and that safety translates to insurance as an example.
Speaker 2 (17:23):
Yeah, just fitting off that safety theme. Can technologies like
scenescape and having those sort of cameras help with worker safety,
say in a factory or warehouse, where they can detect
or even predict, you know, if something's going to go
wrong and actually warn a worker that something's going to happen.
Speaker 3 (17:46):
Yeah. Absolutely.
Speaker 4 (17:47):
Robot interaction is a common one also, so think about
robots with cameras and the cameras and sensors that are
around Mobile World Congress is going on right now, and
I think it was last year. We did a escape
demo there and it was purely an industrial We had
the robotic arms that were moving and they were building something.
And then you have a sensor doesn't even need to
(18:09):
be a camera that has a digital n scen escape
the trip wire, so we know if somebody crosses this point,
then it's a trip wire. So that was the actual demo.
So there was a safety zone and then there was
crossing the safety zone. So if you enter the safety zone,
there could be a warning light to go off. You
don't have to stop anything, but I know someone's in
the safety zone, I know how long they've been in
the safety zone and if they cross past that, then
(18:30):
I know I can start to shut down automatically equipment
if that's the policy that that particular site chooses to use.
So I think can execute another one in a more
constrained environment warehouse that we've seen is where there's actually
a controlled space, so radiation. Actually, the earlier example I
talked about is a real example where there's an area
(18:53):
that it needs to be climate controlled and it literally
has radiations. So they have a radiation sensor both inside
outside and commodity or do you have the equipment on
how long has a person been in this particular space,
and I could set timers and triggers so I know
that they can only be in for so long, and
I can also track that. So that's real time action
and control, and I can also use that for later
(19:15):
analysis and prediction. Maybe I need to change the configuration
of the room, maybe I need to put more signs up,
But you can have real time action and decisions and
also post analysis.
Speaker 2 (19:25):
Yeah, what you said there about the simulation is quite
interesting because you know, as you're talking else, you know
it came back to the gaming side of things, playing
SimCity or roller coaster tycoon being able to sort of
simulate you know, if I put this thing here, is
it going to be dangerous?
Speaker 3 (19:42):
If I put that over there? Does that help the workers?
Does it help with productivity?
Speaker 2 (19:46):
Maybe talk a little bit of some examples of using
real world data to kind of do what if analysis
of various scenarios that management and workers together can can
see mille and potentially improve the workplace.
Speaker 4 (20:03):
So let's take something like a gaming site, I mean
like a football or soccer So clearly those are massive
events with a lot of people, a lot of insurances,
there's safety concerns, there's access to medical professionals that need
to get in and out. So can I take existing
data that has already been captured using existing cameras and
(20:27):
I can actually run simulations on that. I could also
ideally what I want and I need it. If I'm
going to do a digital twin, I need some sort
of digital twin of the environment. The level of depth
is just depending upon the level of analysis that you
want to conduct. Now, what I need is what's the
data that I've been able to collect, because most of
these places they're already going to have some data, even
if it's just camera feed data. I could take that
(20:49):
and actually start to run models on Okay, where are
people congregating? I can actually post camera feed and apply
inference data to that, so I can use the AI
to it. Andy, well, that's a person, and that's an animal.
That's a car over there, And now I can start
to look at Okay, how often are they in these spaces?
Where am I getting congregation? Where am I getting long
(21:10):
killing live? So I can do analysis all on existing data.
Now I can start to reconfigure whatever actions need to
be taken, so all of that can happen before I've
shown up physically at the space.
Speaker 2 (21:24):
Just think about all the personal identifiable information involved in
some of the tasks we're talking about today. Well, the
sheer amount of streaming data coming in from a host
of senses required to implement digital twining, keeping that data
secure is paramount to the future of this industry. I
like to get your thoughts around the whole privacy side
of things, and you know what can be done to
(21:46):
make sure that as individuals we don't feel like where
our privacy is getting invaded.
Speaker 4 (21:52):
That's a very good topic and we thought about that
from the beginning. So one of the ways that we've
defined scenescaper is we primarily work on meta data, and
what metadata simply means is, for instance, we don't do
any facial recognition. I need to know that that's a
person or I need to know In fact, we had
an actual scenario where a customer had a particular area
and they knew people were around, but at night there
(22:13):
were objects and they didn't know what it was. They
were animals, and the model hadn't been trained for animals.
So the model can say, hey, there's something there. I
can't say that it's a deer versus or whatever, but
it's not a human, you know. And so think about
the simplicity of that. Now, I don't have to transmit
every movement, every aspect. I'm only transmitting what's critical to
(22:35):
make the decisions that are needed real time and for
post analysis.
Speaker 2 (22:40):
We talked a little bit about fulfillment centers and warehouses.
I like everyone else's ecomma sites like Amazon Prime. I'm
just wondering if you could maybe paint a picture of
how from the time that I hit that buy now,
but to the time that I get my pair of
socks at my doorstep, take me through how digital twins
(23:02):
could be used. How would a system help that process,
both as an in consumer and also for the business.
Speaker 4 (23:11):
Actually to be honest, One of the first examples that
came to mind is the delivery truck and why location
intelligence is so important. All of us use location based
services today. There was a study I read I think
it was UPS is saving a lot of money per
truck because they realized that location intelligence they were getting more.
(23:33):
Particularly this was when hotels were putting in their address
because they use Amazon Prime to and they're getting these
packages shipped to them. The location data of that hotel
relative to where the truck is coming from, and then
mapping the route were not good routes, so it was
costing the company so much money to get from point
(23:55):
A to point B. So now I can start to
identify where are the hotel. And they discovered this and
they started taking copies which they have.
Speaker 3 (24:02):
This would go.
Speaker 4 (24:03):
We can take copies of the maps, I can start
to locate where am I going. I can start to
figure out the routes. So they're using the twin of
the maps and the data they already have. Think about
they have tons of data on their routes and the
locations and where are they normally congregating, and which truck
should they send, what time should they be They did
all of that analysis to figure out just on the
(24:25):
back end of when I actually dropped the packets off
to you and what makes sense that's saving money for
them completely and again it's location based yep.
Speaker 2 (24:34):
And can you give me an example of how digital
twinning might already be in use for the consumer on
one of these sites such as Amazon Prime all similar.
Speaker 4 (24:43):
I have a chair behind me I just bought, so
now I can use digital twining right now to figure
out exactly where I want this, how does it look?
And they also have those clothing services which are digital twining.
You're the real person, and they have the digital where
you can apply the clothes to you. Yes, I mean again,
these are services that people are using today. But back
to my earlier comment, you're not actually thinking about the technology.
(25:07):
Are taking that now back to work, to your day job. Oh,
I do all of this at home. I should be
applying this to my business and saving money and getting
greater insights of my scene and of my environment. So
those are a couple of examples. Can I give you
a different example. Most people have some sort of ring
doorbell or type of it could be ring, it could
be simply safe whatever.
Speaker 2 (25:28):
Yeah, I'll have that.
Speaker 3 (25:29):
There you go.
Speaker 4 (25:31):
One of the ways we've gone to market is to
make sure what we're doing is standard based and open,
you know, maximum scalability and flexibility. So I have a
ring one of my family members has simply safe. The
challenge is I have the ring camera at the door,
I have another ring camera. I can connect them. I
can see if something's walking by. That's great. It's not
(25:53):
very open. I'm somewhat siloed. What we've had someone do
with scenescape is they use scenescape. First of all, you
don't need to go through the so they're not paying anybody. Okay,
if you want to use the cloud, you can, but
you do not have to use the cloud. Everything can
be edge based, and by edge based, think about the
edge again. At home, where's the data generated? That's my edge.
So in this case, your edge is your home. So
(26:13):
my home, I already have a computer, and I already
have one or two cameras. But there's a particular type
of camera I want for the front yard, which is
totally different than the camera I want indoors, which is
totally different than the camera I want, so three totally
different brands. Well it's called multi camera multi brand from
that sense. So and by the way, maybe I want
a heat sensor or something like in a particularly air
in the backyard because I don't know if there's something overheating.
(26:35):
So now I can add all different type of brand
sensors and I can connect that into scenescape. And now
scenescape party has AI. So I've used different brands. I've
used my own computer, so I have standard based connectivity.
And because of standard based connectivity, I can connect it
to my phone. Every phone app now it's very easy
to connect to it and get alerts. So now I
(26:55):
can start to use the existing AI tools that are
in Scenescape. But there are so many applications out there.
With scenescape, you can integrate it with other applications. So
there was one person that used it to identify the
difference between a car coming in the driveway versus a
postal truck that goes by and stops, and they set
an alert. So whenever the postal truck comes and stops
(27:18):
for a few minutes, the alert goes on the phone.
He never has to look at a camera. He knows
when he gets that alert mails here. You can't do
that with RING, you can't do that with these other applications.
So that's a common use case that people know today
where standard digital twinting technology with AI, standard based communication,
and standard computing technologies can all be used to enable
(27:42):
use cases that we use every day.
Speaker 2 (27:44):
Yep, we just have time for one more question. I
would like to get your number one. I guess area
of excitement for digital twins for.
Speaker 4 (27:57):
Me is healthcare. What I want to see in my lifetime,
and we have the technology to do it. In fact,
we've have a few use cases with scene skates where
we're working with the medical community. I want the digital
twin of my health. I want for me the person
all think about all the data, all the medical records.
Speaker 3 (28:14):
First of all, it's hard enough keeping all your medical
records together.
Speaker 4 (28:16):
So not only my medical records, but the medications I've taken,
any reactions I've had. You have so much data from
blood work and positive reactions, negative reactions to medications, exercises
I've done that may have improved weight or blood pressure.
So as I grow and as all of us grow
and age, I should say yes, I want all of
(28:40):
that history to follow my DNA, my person to maximize healthcare.
I am the ultimate digital twin that I want. I
don't care about an avatar that's fun and fancy. I
want something that helps to improve my quality of life.
That's what I want out of interest. Have you seen
any companies or businesses looking into this. I did meet
(29:00):
a company or CEO of a company that's working on
the medical record side where they're trying to tie all
of that to the person so that can follow them,
so that now physicians and healthcare workers can have all that.
So it's a startup. It seems to be much more
challenging to get this done than you think it would be.
But we've engaged with some companies and hospitals that are
(29:22):
making their hospital smart. You can see some areas called
the smart operating room. That's a particular area in a
hospital that's obviously critical. I mean, you think about something
as basic as we're in the operating room, we're starting
the operation. I have twelve high value instruments on my right.
Those twelve high value instruments need to be there when
(29:42):
I finished, because if they're not there, there's a very
bad place they could be. Ye, that's right, and that
is a real example that I know personally somebody like
that that's happened to and they've had to go back
and get one of those instruments. So when you think
about the seriousness of the operating room, and that's before
you even get into intrusive sensors and I mean, you know,
what's the blood pressure and et cetera. Yes, you never
(30:04):
go to a hospital, are to a healthcare professional and
they take your blood pressure and that's it and you're good.
Speaker 3 (30:09):
Then they start talking to you.
Speaker 4 (30:10):
No, we don't even think about the fact that we
take blood pressure, we take temperature, we take weight, sometimes
we take blood. So we're already experiencing a multi modal
environment to maximize our health, but we don't always think
about that when we bring that to work. So now
I need a temperature sensor, I need light, I need
lie dar, I need cameras, I need different brands. I
need to apply intelligence to that. So now I can
(30:32):
perceive my space and my environment, I can understand it
with analysis and AI and make sense of it to
make decisions, and then I can also do prediction based
on that. So what should happen in the future.
Speaker 2 (30:43):
That's great, Tony. I think we will leave it on
that note. Thanks so much, No.
Speaker 3 (30:48):
Thank you, this was fun, really enjoyed it.
Speaker 2 (30:52):
My deepest thanks to Tony Franklin for sharing his equities
with us. Today's chat about digital twins really our eyes
to the incredible potential. It's like stepping into a simulation
game where you can tweak maintenance schedules, production lines, and
even play around with the interaction between workers and machinery
using real world data. Yes, I'm letting my inner geek
(31:14):
shine through here, but the idea of managing a supply
chain with the ears of playing SimCity seems pretty cool
to me. Tony's closing thoughts on the future of healthcare
and the possibility of creating a human digital twin we're
particularly striking. Imagine having a clone of yourself in a sense.
I mean, we're already wearing watches that monitor heart rate,
(31:35):
physical activity, sleep quality, plus a range of other biometric data.
It's not too far fetched to dream about a future
where digital twins can forecast our health outcomes based on
our DNA, diet, and exercise. It's an interesting idea and
I'm looking forward to seeing where this technology takes us,
(31:55):
and lucky for us, we'll get a chance to explore
this further on our next episode Tuesday, April twenty third,
on technically Speaking, an Intel podcast, we'll be learning about
some of the revolutionary implementations of AI in the healthcare
space with team members from Intel and Siemens Healthineers See
you then. Technically Speaking was produced by Ruby Studio from
(32:23):
iHeartRadio in partnership with Intel and hosted by me Graham Class.
Our executive producer is Molly Sosher, Our EP of Post
Production is James Foster, and our supervising producer is Nikia Swinton.
This episode was edited by Sierra Spreen and was written
by Molly Sosher and Nick Firshall. Where do world changing
(32:51):
ideas get their start? At Intel? It starts with real solutions,
and real solutions start with exceptional engineering, the quantum computing,
the next generation of AI experts, the renewable energy grid,
liquid cooling, data centers, early diagnosis for cancer, water restoration,
and even farmland protection. The examples are countless, the impacts
(33:12):
are endless, but the foundation is always the same. It
starts with Intel join us in redefining what's achievable through
the power of AI. Learn more at intel dot com.
Slash stories
TechStuff News
Host
Jonathan Strickland
Popular Podcasts
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations.
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.
The Nikki Glaser Podcast
Every week comedian and infamous roaster Nikki Glaser provides a fun, fast-paced, and brutally honest look into current pop-culture and her own personal life.