Original subtitles

thank you so much Iris and welcome also from my side

it's an honor and also really a pleasure

to have the opportunity to speak to you

um today at the GTC

and also that there is

quite some curiosity in the audience

uh for this topic um the time is a little bit tight

um I could maybe speak 2 hours about this one

but anyhow before we start I would like to

give you a few minute or minutes or a moment

to think about

what we heard in the last days

I guess many of you also participated in the sessions

there are many innovations many applications

state of the art technologies and things

which we want to apply in the future

in the manufacturing business

but if we think about this one

and series production large scale series production

there are many

let's say

problems or challenges how to apply those things

so if you take a moment and think

about what in your industry

maybe at your company

is the biggest

challenge for this one

maybe you found something

maybe some of you work in the automotive industry

and at least from our perspective

large scale application is one of the biggest things

simulations are already there since decades

AI to a certain extent already

we are speaking about since years

gen AI is the new platform so these are things

which are not completely new

still in many industries

we are not doing them yet from end to end in a really

sophisticated and daily business

today in the next 23 minutes I would like to show you

one way how we believe this is possible

this is for sure not the way but is a possibility

so before

let me put this shortly please into perspective

where are we coming from

why are we doing this who are we

as mentioned I work on before Continental

and we are an automotive supplier

in total with 200,000 employees all around the world

and specifically I'm working in the business area user

experience and we like to say

user experience is the new horsepower

some years ago

when you bought a new car most probably

you looked at the engine

maybe the fuel consumption things like this

today what many people

look at when they buy a new car

is how connected is the car

what kind of features does it have

what kind of automated driving possibility

these are our products and what you see here

is a sophisticated product called pillar to pillar

this is our one of our core products

but for sure also head up this place

and full digital clusters

this is important to say because

in Germany at least many

many people when you say I'm working for Continental

they think oh you do tires

no we are doing much more than tires

basically everything

a part of the car which you can think of

we are producing besides an engine

and the chassis

the rest somewhere we have business for this as well

and this is especially important because

in this electronic manufacturing environment

we have challenges which for example tire industry

does not have

most of our OAMs produce a new car vehicle or a new

model variant every 2 to 3 years this means for us

new manufacturing pros as as new technologies

in order to build this

a display is no longer just hardware

the most complexity of our product is in the software

so also a transition is happening

that we are becoming

more and more from a hardware company

to a software company

and we need to adjust on this one

so this means putting all into perspective

there are different

needs in our factories

needs on the shop floor

and as I was mentioned I'm working in the industrial

engineering area

so everything what I'm telling you about today

is focusing on shop floor is on operations and on

let's say in general producing parts topic

real time data for sure

is something which is already on their

since years as well

still many companies have difficulties with collecting

really end to end

everywhere from

anywhere at any time

all their real time

data and if they are able to collect them

they are maybe not yet fully able to fully utilize them

so these are challenges

especially in the automotive industry

which are now some general statements but I guess

if you are working in this industry somehow

you can relate to this

out of this

we have certain requirements and these requirements

bring us now from a more

reactive manufacturing approach to a

predictive manufacturing approach

this leads to some of the biggest things

which I'm trying to present to you today

is that we need to start speaking about intuitivity

how can we transfer all of these AI things simulations

into our business into the daily work especially

when our factories have to produce every single day

parts this is what they are doing

the right part in the right amount of time

at the right quality to the right cost

and

this we need to bring also to that engineering people

so the best simulations will not help us

if we are not able to transfer them

into business into the real doing

and this is what we are trying to achieve

so we need

kind of a digital factory platform so that we have

not only a digital twin

not only a process

not only one stage which is there

connected to real time

we want to have a full factory connected

I want to be able to sit wherever I am

in Germany in Frankfurt theoretically on the moon

I want to be able to collect

connect in real time

to the factory and see what is going on there

and this is a real shop floor use case

for shop floor people

for focus factory people for managers

who are there to operate the shop floor

we found a solution

for this topic

and this is the one I would like to present to you

and for this

we have a very simplified overview of an architecture

so there are different

things from where we are collecting the data on the

shop floor there are factory data

where we also have really sophisticated systems

like in general and Mas system and manufacturing

execution system where

a lot of holistic data is in the quality data

product data and so on and so forth

we have product a data

coming from our R amp

d department

and that's what I say when I speak about end to end

really following an entire product lifecycle

from a quote process until the product development

and then the development of

how I need to produce this product

so

these are both things which we want to consider here

also then after serious production the aftermarket

and then for sure

and this is the biggest thing

I guess here in the center

so on the left side still sorry

the line and the station data

and here we are speaking about

real time data

from process and machine

torque pressure temperature

you name it all of these kind of things

then go into let's say a cloud based data lake

in which we are combining

all kinds of data so that we have basically

a baseline set

this is what we try to achieve here

because without the data

a simulation or

an AI model for sure will not be able to operate

in the end and this is now where we are using Omniverse

is as a kind of an aggregator or a platform which

allows us for the very first time

to have multi

platform and cross

platform cross discipline collaboration also

and we are using different cornerstones of

the Omniverse and for sure

we are also investigating right now into

some of the new things which have just recently been

mentioned

but how does this come to life

and how is this really going into an engineering

application where

engineers can work with the tools they know

there are solutions on the market also from some big

German simulation providers as well

which offer already something like this

an entire toolbox from end to end

however if you go into such a solution

you take a product from the shelf

and you cannot adjust it

you are

related or reliable on one supplier

you are dependent basically

and also this is the most important

aspect from our side

you have an engineering team

across the world who have own tools

own processes own workflows so it's almost impossible

to

switch from one day to another without losing all your

overall equipment effectiveness

from one tool

to an entire tool landscape from one specific supplier

so this is the big advantage here

that we are basically open source we can

just program our own plug ins

with an omniverse

we are not using this so much for this

nice rendering and photorealism

this is all good and nice

but this is for us at least right now

the cherry on the cake

what is for us much

much more interesting is the aspect of USD

of the compiler that basically

any engineer no matter with what tool he is working

can

transform his data into USD

and then interexchange data

from one discipline to another

from one tool to another

and by this you are reaching really an end to end

approach from product design into series production

but nice lights are good and for sure I could

maybe sell you here the world if I'm smart enough

but I'm not a salesman

I'm an engineer

but videos and pictures and

use cases I guess explain much

much more than just slides

so I'm really happy to introduce to you what we call

the controversy

so the naming

the naming is most probably not

you don't need to be rocket signs for

metaverse omniverse contiverse

but omniverse it is

so what is contiverse

and this is a video out of our live application

so what everything what you see here is live

so it's live on

our service basically

even if you would have the link and the continental

email address

you would be able

directly today to access this one

so what we see here is a way

how we want to bring

Omniverse with a graphical user interface to engineers

and it starts with you log in

and you have your Avatar

we see here questionize one of my employees

and what does questiona have here

now he has certain activities in the middle

things he want to do

he can achieve rewards he can do participate in audits

so he has certain roles now

which he can select out in his daily work

depending on the tasks which he has

he has

reviewer mode a planner mode a visitor mode depending

on what exactly is today

on his agenda

out of this

he is now basically able to select and

in our controverse

depending on the role you are selecting

you have now the access to different features

different functionalities

different applications

if we go next and this is now bringing this

idea to live what I mentioned

we see a world map

easy like Google Earth I guess everybody knows it

here you see our current

producing factory network and the intention is

as I mentioned before

from anywhere and anywhere

you can access a digital twin factory

we have the pilot for our

new mega plant in Novisat Novisat is just

went life

into series production one and a half years ago

so the intention is you click

you see the most important information of the location

you see how many employees you see what products

you see what is the volume you see what is the overall

effectiveness of the plant

so you have all of the information of one factory

at hand in one side

you log in and you see and this is nothing new for you

it is basically an industrial metaverse platform

you get and once again

the focus from our side is not so much on the visuals

the focus is on the functionalities to have

an interface which is accepted and intuitive

to use from engineering

so we have an Avatar system so

directly integrated in the Omniverse

so this means you can meet with colleagues

you can come together

you can join one session

you can group up

you can interact with this and this for sure

allows not only

shop floor and engineering use cases this is

the tip of the iceberg

this allows use cases for basically any discipline

but we want to focus on the shop floor

and we want to start with a virtual Gemba

on digital twins

Gemba for those ones who are um

familiar with the shoplore environment

Gemba as a typical term comes from the Japanese

means basically go and see so with your own eyes

and virtually

we can do this and this was

especially a thing in Cojona

which starts now to allow us bringing

experts remotely

from basically anywhere in the world together

and this is what you see here now

our factory in Novisat

we were

able now also not only to have one station or one line

in the omniverse

we have the entire factory

and everything so this is fully scaled

and

some of these stations and some of these lines

which you see here

are already connected via opcua

directly to the network so these are not all

just moving based on a simulation

they are moving based on real time

but this we are not doing it for everything

so not everything is here fully connected yet

what we are seeing

but some of the stations as pilots we already connected

to the real world

so full digital twins when there is a line stop

this system will stop

when the line is producing 1,000 parts an hour

this system will produce 1,000 parts an hour

we see here also

something which we cannot do in the real life

with one click access to performance data

you click a line you click a station

you are able to access the performance and not with VR

VR is nice they are

for sure AR and VR use cases

but I don't need to go in the

VR to walk through the line

this is a nice gimmick

but if I do this on my computer

like in a computer game

it is much more intuitive for an engineer

much more accessible

you see one of our own dashboards

you see Powerbi integration

so basically now we have the chance

really to

plug and play any tool you are using also in such a

environment because it is

with the different plugins which you have

basically

completely scalable and completely adjustable

what we have developed with our partner t systems

is also this Avatar system as I mentioned

you have to now have the chance to invite users

you see who is online

who is currently in the contiverse platform

and you can invite somebody just with a click

and then there's a pop up page

do you want to accept my invite

and together with the person you are able now to go

to the shop floor virtually

check certain stations where are the problem shooters

where are the troubleshooters

you could now click again on the line

see what is the data

and you also have the possibility

but this is right now still the first version

to get feedback what do we have to change

you can send out of this system alerts or pings

to the experts which are maybe not yet

able to join you in person

so this is the first advantage

the second advantage or second

use case which I would like to introduce to you

is editing new lines

with Omniverse

we have achieved as I mentioned before

a direct connection of our engineering tools

with the platform

what you see here is a 1 to 1 correlation

on the right side our engineering tool

on the left side omniverse

so everything what we are doing on the right side

what an engineer is doing

in real time

is accessible for anybody who is in the contriverse

imagine use cases like you are

planning your shop floor you are planning a new line

you need to change the layout that these are

easy things layout topics

but if you want to change a process

a robot movement

we heard a lot about this yesterday and

also in the keynote

digital twins

when you want to do an offline programming

and you want to validate this

changing the robot sequence

clicking upload goes and

directly transferred into the tool

this is

where a real application use case comes out of it

creating new line concept in a collaborative mode is

another

use case which I would like to introduce to you

this is

showing a capability

how open omniverses how open conteverses

because we have an internal web application

we call this internally ESRQ

imagine ESRQ like Amazon

but you cannot buy books or TVs or whatever

you can buy equipments

so here we have a standard equipment library

it is our platform

you see the equipment and you can create

often out of the different equipments

drag and drop align

this is what we are doing here in our

web based application

engineers

planners persons in the quote

just

drag and drop a new manufacturing line in the system

and the basic 2d lay

it seems simplified but the basic 2d layout

and this is now what this

important thing is

the basic 2d layout what we are creating here

with two clicks

we are able to upload into the omniverse

so this means

somebody from a sales team

somebody from an engineering team who

maybe is not a simulation expert

can say okay this is how the line should look like here

clicks on transfer

and somebody who is

able to use a simulation tool suddenly

already

has the basic line concept and now can do the fine

tuning and the fine planning

this for sure

especially when we are speaking about engineering ours

is a massive reduction

for sure we have the standard library

from our ESRQ

platform also connected to the simulation tool

so that it's really one workstream

and this I can really not emphasize enough

the main

importance of this application

is not the avatars is not the

surrounding it's not

the possibility to connect it with AR or VR

glasses the main advantage is

the workflow behind

that for the first time

at least for us

the simulations AI

data is coming together

in a tool which can be very intuitively and

easily applied

from engineering people

virtual improvement workshops is our third use case

um is basically

something which we are doing in our business daily

that we are reviewing lines which are either

still in the concept

or also that we are reviewing lines

which are right now in the

series production already

we meet we go there

we go to the troubleshooter stations we try to check

the data why are this

screwing station or whatever

why does this have a problem

this is with the high fidelity

which is for sure nice to have it

also good we can go into

areas which normally a human being could not be going

if the line is right now producing

in serious production

but only you see the issue when it's in serious

you cannot put your head inside of the cabinet

and check what the robot is doing here you can

and lastly or not lastly but one

other point which comes hand in hand to this

is with a few clicks you can start comparing lines

with each other

we have here a selector

in which all the lines from Novisat are in

we just select those

for sure they are now a little bit randomized

at different customer projects

we have at the plant

you compare and you get an easy overview

on what is the line performance

what is the layout of the line

how much top floor space

what is the OEE

how many stations do we have in there

what is the individual performance of every station

what is the output

how much OK parts versus how much NOK parts

what was the plan

so all of these things

and this is especially for us important

because we try to

standardize our lines as much as possible

so we try even though for different customers

to standardize line concepts

and just to exchange the equipment

so then we can start seeing

which standard concept is the best one

and lastly

the detailed planning

also fully connected to Omniverse means that

if you want to make

changes or if you want to test a different scenario

you have it here

and there is no versioning anymore there is no

oh sorry sorry

somebody is right now in vacation

but has the file on his

personal desktop computer

I'm over exaggerating this now but I think it's

everyday business

and here everything is in the cloud

everything is accessible

from anybody at any time from anywhere

yeah as I mentioned contiverse for us is an ecosystem

and this ecosystem is

compiled out of many many many use cases

and this is something where I'm also quite proud of

we are no experts for omniverset

we are no experts for AI it

but we are experts for our production

we know what problems we have

we know how we want to solve them

and we are the ones who have the ideas how we can

bring AI simulations together

and for that reason the four use cases

I just showed you is for us just the beginning

there are many more

maybe next year the GTC there is

another possibility for us to present even more

last but not least

for sure I also would like to speak a little bit about

where do we see the biggest advantages of this one

for sure it comes mainly down

to cost to invest and to engineering ours

as I mentioned the platform is about

to focus on a workflow

so when we see the biggest benefits

it's firstly and most obviously

that we are able to save engineering hours

along the entire planning process

from start to end

but once we are having more and more data

and also AI tools connected for sure

we are expecting

operational performance to increase

cost for shop floor to increase

energy cost to decrease in this case

all of these kind of benefits we are

expecting and we have also made some estimations

with our experts together

and with this I would like to come to an end

and

I hope and I'm really curious

about your questions maybe you have some comments

but I would also be very happy

if you use the QR code and we connect

I'm here still today tomorrow full day

so I would be really open

for an exchange on any of the topic thank you very much

thank you Benjamin for this awesome presentation

now let s open up for a quick Q amp a any questions

in the room

just it's a peripheral question actually

the German automotive industry is building

a certain data interchange standard that

seamlessly connects all vendors

pretty much all exchange of data takes place seamlessly

um I was curious how that place alongside

Omniverse I mean is that

standard built into Omniverse today

can you use it

can vendors plug into your Omniverse or controverse and

are you specifically referring to Catena X

Catena X exactly okay

I guess so

yes so absolutely so Catena X is there Catena X is just

a short Catena X is

an automotive

standard basically

um funded also by the German government

and the intention is

that with Cartena

X there is a seamless data exchange from

basically all suppliers vendors Oems

so this is what Cartena X is in a nutshell about

so and the answer is yes

but um we want to open this

uh controverse in the future also to OMS uh to

our suppliers

um but for sure we have to think about cybersecurity IP

not everybody can do everything in here for sure

and but the intention really is that

there is a seamless data

exchange possible

from our equipment producers

towards us and then towards the OEMA

because

we are thinking about demands of forecasting or supply

chain then for sure there are

thousands of use cases where such a system would be

amazing to come into places

thank you for your question next one

yeah does

whatever happens on the plant floor and the controverse

ever get out of sync

and do you have process to prevent that from happening

or making sure it gets back in sync on

like the same state

so so you're referring um

basically when there is any

out of sync from real time data happening

yes exactly so

for sure there

needs to be some kind of security strategy

some backup plan happening

what happens if

then there needs to be a shutdown

um but since we did not yet fully roll this out

to a large scale we are currently

doing this for one factory

and as I mentioned before

this real time data connection is we have

established I think in total for

two lines fully

and one or two additional stations

this we are currently monitoring still there

so there is no

possibility that somebody is now accessing and

yet there is a standstill or whatever happening

because currently it's still on the monitoring process

but this definitely is on our radio

so there is

right now not a technical solution yet implemented

to answer your question

thank you for the question

any other questions

yeah

yes

yes exactly

this would be this would be the next step for sure

our factories everywhere have sensors

safety sensors

not exactly sure late laser

curtains I think light curtains is the right word word

yes this we have but it's not yet in there

I mean you can go from

large scale to the last tiny bit of sensor right

but the intention really is that this

that this platform is scalable to this level

so that you also have a safety stop

but maybe you don't even want this

because if I walk there as an Avatar

maybe I don't want

the robot to stop maybe I want to exactly see

from an ankle which I normally are not able to see

so it depends a little bit on the use case

which you have

and how much

how much time and

how many people did work on the on the project

from from the initial IDS to

the demonstration that we have

so to be completely Frank

contiverse is an idea which we already have since

a year one and a half

it took us quite some time

to find out what the right platform for this is

there are many many many

let's say industrial metaverse platforms around

which we tried

until we finally reached Omniverse

and what you saw today

taking now the idea itself so that just the technical

implementation

is from October

until today and the amount of people

in my team we have

three people working on this but also not full time

and as we are collaborating with R amp d we have here

the main development team also behind I would say

8 people

6 to 8 people but not full time

so maybe F for FT

something like this

thank you for your question any other questions

cool that's all we have today with Benjamin

thank you so much for attending today's session

feel free to thank you for joining this session

please remember to fill out the session

survey in the GTC app

for a chance to win a 50 dollar gift card

if you are staying in the room for the next session

please remain in your seat

and have your badge ready to be scanned by our team

which one your power point

um it will be

as far as I know the presentation was recorded

and after 72

hours or something

it will be uploaded to whatever Nvidia

we can check it with Iris regulator