AWS re:Invent 2021 - {New Launch} Introducing AWS IoT TwinMaker

Aug 16, 2023

AWS re:Invent 2021 - {New Launch} Introducing AWS IoT TwinMaker

AWS IoT TwinMaker makes it faster and easier for developers to create and use digital twins of real-world systems like buildings, factories, industrial equipment, and production lines. This session reviews key features and use cases and shows a demo of the new service.

Learn more about re:Invent 2021 at https://bit.ly/3IvOLtK

Subscribe:
More AWS videos http://bit.ly/2O3zS75
More AWS events videos http://bit.ly/316g9t4

ABOUT AWS
Amazon Web Services (AWS) hosts events, both online and in-person, bringing the cloud computing community together to connect, collaborate, and learn from AWS experts.

AWS is the world’s most comprehensive and broadly adopted cloud platform, offering over 200 fully featured services from data centers globally. Millions of customers—including the fastest-growing startups, largest enterprises, and leading government agencies—are using AWS to lower costs, become more agile, and innovate faster.

#AWS #AmazonWebServices #CloudComputing

Content

0.82 -> Good afternoon, everybody.

2.72 -> Welcome to our session.

4.98 -> This is intro to AWS IoT TwinMaker.

9.24 -> It's a new service that was announced yesterday

11.6 -> in Adam's keynote.

12.92 -> So if you watched the keynote and you're curious

15.97 -> about what TwinMaker is,

18.48 -> or you're thinking about applying digital twins

21.77 -> to your operations,

23.65 -> or you're already building digital twins and wanna learn

26.94 -> how the service can help you get there faster,

30.7 -> then this is the session for you.

32.93 -> My name is Saras.

34.26 -> I am a principal product manager in AWS IoT.

37.94 -> And with me today is Dane Loughlin from Invista

41.6 -> who's here to talk to us about how industry leaders,

44.37 -> such as themselves are applying digital twins today.

48.18 -> We're really excited to share with you what we learned

51.08 -> in this space over this experience,

53.42 -> and hopefully inspire you to think

56.62 -> about how you can apply digital twins in your businesses.

62.67 -> Thank you for being here and with that, let's get started.

68.96 -> On the agenda today we'll cover a brief background

72.38 -> on what digital twins are and how they're used.

76.23 -> We'll talk about how digital twins are built

79.97 -> and what makes it so challenging.

83.1 -> We'll take a deeper look into TwinMaker's capabilities,

88.26 -> and then I'll hand it off to Dane to talk to us

91.27 -> about their motivation

93.73 -> and experience building a connected worker tool

98.19 -> using this new service.

101.18 -> Okay, so what are digital twins?

106.29 -> There's a lot of interpretations out there,

108.54 -> but simply put digital twins are virtual representations

113.43 -> of physical objects and processes

117.2 -> that very closely mimic the state and behavior

122.12 -> of the physical counterpart and are built

125.16 -> to solve specific business challenges.

128.65 -> Customers are using digital twins to get better visibility

132.84 -> into their operations,

134.89 -> to get a deeper understanding about their environments

137.79 -> and looking for ways to improve and bring improvements

142.08 -> into the environment and even using analytics

145.67 -> on the digital twin to generate insights

148.23 -> that are then applied to the physical world

151.07 -> to drive those improvements.

153.73 -> But digital twins are not really a new concept.

157.03 -> They've been around for a long time,

159.62 -> but what makes them relevant now is customers

163.58 -> across industries are focusing so much

166.86 -> on digital transformation,

169.27 -> as well as advanced technologies like graph databases,

174.93 -> immersive visualizations, broad adoption of IoT,

178.6 -> machine learning and cloud computing

180.78 -> in general has made digital twins, not just possible,

185.64 -> but within reach for all our customers.

191.44 -> Going back to the solving of specific business challenges,

196.48 -> nobody wants to create digital twins

198.98 -> for the sake of creating digital twins.

200.98 -> It's really hard.

202.91 -> What the first question we have to ask is,

205.65 -> what is the outcome that you are working towards?

209.27 -> That's the only way to define the problem

213.2 -> and then solve that problem.

216.23 -> So if you take the business outcome as a given,

220.36 -> now you can start to think about what data

224.11 -> or input needs to go into it.

227.05 -> You can start to think

227.95 -> about what insights you wanna generate from that data,

232.23 -> as well as what's an intermediate business outcome

235.25 -> that gets you part of the way there?

237.61 -> When you answer these questions,

239.13 -> it gives you a better understanding

241.47 -> of what your digital twin needs to look like.

244.74 -> How are you going to use it

245.98 -> and what improvements it's going to drive?

251.27 -> So we spoke to a lot of customers across industries

254.36 -> and they shared with us a lot of interesting

257.91 -> and innovative use cases.

260.36 -> What we found was that a majority of customers are focused

264.37 -> on applying digital twins to transform their operations.

269.34 -> I'll give you two examples.

271.78 -> When an operator is trying to root cause

274.74 -> and correct operational issues in manufacturing,

279.32 -> they're slowed down by having to constantly context switch

282.74 -> between different applications that have the data

285.42 -> that they need to solve the problem.

288.03 -> They're pulling up asset information

290.09 -> with a specific asset ID.

291.81 -> Then they're trying to pull up quality

293.57 -> or maintenance history data with probably an asset ID

297.37 -> that does not map to the other applications.

299.83 -> They're switching between all these different dashboards

302.57 -> and this kind of context switching

304.15 -> and different applications slows down resolution

307.73 -> of the issue and can lead to, you know,

311.45 -> equipment failures or downtime.

314.32 -> But what they really need is the ability

317.46 -> to visualize the data that they need to get the job done

322.58 -> in the context of the environment

324.28 -> that they actually have to apply this to.

327.04 -> And they wanna do this using a comprehensive application

330.98 -> that is essentially a single pane of glass,

333.7 -> giving them real time 3D view into the facility

337.37 -> so they can quickly pinpoint and address

340.27 -> those equipment and process anomalies

343.75 -> and reduce the time to resolution.

346.76 -> It also enables remote experts to be able

349.51 -> to have a comprehensive understanding of their operations,

353.97 -> making it more effective for them to do their jobs.

358.07 -> As another example, customers in real estate are

363.08 -> hyper-focused on improving occupant health,

367.8 -> safety, and comfort.

370.07 -> And the way they wanna do that is by collecting data

374.49 -> about temperature, occupancy data,

378.27 -> energy consumption data, air quality data,

381.2 -> and visualize that in the context of their environment

384.94 -> to see what are areas of improvements

387.32 -> that they can focus on.

389.87 -> Also allows them to visualize how energy is being consumed

393.86 -> in the building at a granular level.

395.9 -> So at like a floor level or an individual room level

399.25 -> to see like what are some optimizations

401.53 -> that they can apply to start to think about, you know,

404.36 -> reducing their carbon footprint.

406.78 -> But these are just two examples of a vast array

409.98 -> of use cases that digital twins can be applied to.

412.83 -> And we find a lot of customers want to start here.

415.7 -> They want to connect to data sources

417.84 -> from all these different sources,

419.63 -> bring it all together, visualize it,

421.98 -> and then start to understand

423.68 -> what are improvements I can make,

425.59 -> what are areas that are ripe for innovation

429.41 -> and will drive improvements in their businesses.

437.19 -> So let's talk a little bit

438.71 -> about how do you make these digital twins?

442.33 -> So customers and companies are collecting

444.83 -> and generating so much data, not just data in volume,

449.59 -> but different types of data.

452.39 -> You see here, there is IoT and process data

456.12 -> that's being collected from IoT sensors and devices

459.11 -> and equipment that tell you the current state

462.16 -> of these machines.

465.29 -> There is visual data.

467.961 -> The as designed context from CAD files

472.44 -> or as built context from point cloud scans,

476.1 -> there's geospatial data,

477.93 -> there's modeling data as well as video data

482.16 -> from cameras that are deployed in these facilities.

484.98 -> These are all very different kinds of data.

487.9 -> But when you're thinking about building digital twins,

490.77 -> you have to bring all of these pieces together,

493.62 -> depending on your use case,

494.75 -> you might need all of them or a subset of them,

496.53 -> but you do have to bring them together in and codify them

499.99 -> into a virtual model

502.12 -> that represents the physical environment

505.17 -> where these data sources are relevant,

507.27 -> how they're related to each other,

509.17 -> and what are the various elements in that environment.

512.45 -> And you wanna use that to drive insights.

514.96 -> And once you have that digital twin,

517.07 -> you wanna embed those into end user applications.

522.32 -> These are the applications that end users use

526.38 -> to consume the data and insights

529.77 -> from your digital twin and use that

532.27 -> to do their day-to-day jobs

534.39 -> and find those areas of improvement to improve operations.

541.429 -> What we find though, is that there's a layer of shared data

547.7 -> and resources that is generated

549.92 -> that truly forms the foundation of your digital twins.

553.71 -> And if you think about a typical asset going

556.58 -> through its life cycle from design build operate,

561.17 -> there is different types of data

563 -> and resources being generated and collected.

565.84 -> For example, in design,

567.1 -> you're getting your CAD files,

569.55 -> your builder material.

570.92 -> In your build you have other vendor information.

574.75 -> When you get to operations,

576.05 -> you have maintenance, IoT data,

580.76 -> all of this data truly forms the foundation

583.56 -> of what needs to go into that digital twin.

586.18 -> And that's why we believe AWS is the best place for you

589.09 -> to be building your digital twins because of the breadth

591.81 -> of solution that we have available from edge solutions,

595.3 -> for data collection and ingestion, purpose-built databases

599.71 -> to store different kinds of data, advanced analytics,

602.933 -> serverless compute resources available to you.

607.03 -> So essentially we wanna bring the power of AWS's

611.03 -> entire portfolio of services to bear

613.34 -> on your digital twin use cases.

615.38 -> And for that data is a key enabler.

621.04 -> When we talk to customers,

623.36 -> they tell us digital twins are really hard to build,

625.6 -> and we believe them.

627.41 -> Building and managing digital twins, it's complex,

630.91 -> it's time consuming, it's complicated, it's costly.

634.8 -> And it requires specialized skills from developers

638.68 -> that not all customers have.

641.35 -> The three main key pain points that customers point

645.41 -> to us is I have so many different data sources,

649.87 -> these data sources hold important context

652.7 -> for my digital twins,

654.14 -> but there is no seemingly easy way

656.8 -> to connect the dots between them.

659.74 -> The second challenge they describe to us is,

662.89 -> modeling physical systems is very complex

666.78 -> because physical systems are complex in nature.

669.87 -> And being able to do that,

673.1 -> that tracks in the assets lifetime gets even harder

677.11 -> because you have to now think about version control.

680.76 -> And third, digital twins are used by the end users,

684.8 -> and you have to make these visualizations effective

687.55 -> for the end user and it has to be applicable

690.57 -> for the use case and for the persona

693.13 -> that's using the digital twin.

695.22 -> And in a way that makes their job easier.

697.61 -> So should not have information overload,

699.96 -> but also should not be context switching.

702.12 -> So it needs to be very specific to the challenge

705.05 -> that customers are trying to solve.

709.54 -> So that is why we're very excited

711.5 -> to announce in preview AWS IoT, TwinMaker.

715.13 -> These challenges that our customers have told us

717.41 -> over and over again, was the motivation

719.33 -> for us to build this service.

721.53 -> So we're really excited that we're now able

724.86 -> to make this available to all our customers.

728.23 -> What is this service?

730.6 -> It's a new AWS IoT service that allows you

734.03 -> to make digital twins of real world systems

737.97 -> faster and easier than ever before,

739.99 -> and use them to optimize your operations.

745.29 -> The three key benefits of the service, one,

748.33 -> it lets you access data where it lives

750.93 -> from various different sources

752.92 -> without having to re ingest the data,

755.81 -> without having to move it all into a single repository,

759.41 -> so truly the data can live where it is.

762.75 -> Second, it allows you to

764.87 -> accurately model your complex physical environments

768.76 -> into a knowledge graph that understands

771.75 -> all the different elements in that environment,

774.95 -> the data sources behind those elements that are relevant.

778.57 -> What are the relationships between them?

781.03 -> What are the interactions that they have?

783.08 -> So you can start to create a virtual digital version

787.58 -> of that physical environment.

789.72 -> And then third, creating immersive 3D views.

793.61 -> A lot of the customers that we spoke to are really focused

797.47 -> on bringing those 3D next gen experiences to their end users

802.33 -> because it's been proven that visualizing data

805.84 -> in its natural context makes it easy

808.34 -> for operators to understand

810.42 -> and comprehend that the data and insights

812.65 -> that they are trying to use to get their jobs done.

816.23 -> So those are the three key benefits of this service.

820.08 -> This is a very visual concept.

822.62 -> So I wanna play a quick video so we can get

825.47 -> on the same page of what kinds

827.17 -> of applications I'm talking about.

830.12 -> So this is a cookie factory demo that we've built.

833.38 -> What you're seeing here is you can easily bring

836.67 -> together data, sensor data, equipment data,

840.76 -> video feeds from cameras and other contexts

844.05 -> into a consolidated digital version of your operations.

849.15 -> You can create immersive 3D visualizations

851.97 -> off your environments, showing real-time data video,

855.88 -> and all the contexts that's been modeled

858.57 -> to help your operators make better decisions

861.35 -> and make them quickly.

865.78 -> So how does it work? How did we make that?

867.56 -> There are four key pillars and I'll talk

870.66 -> about all four of them in the next few slides.

873.61 -> It's Data Connectors, Model Builder,

876.49 -> Scene Composer and Application Toolkit.

881.78 -> We'll start with Data Connectors.

884.49 -> So data connectors from TwinMaker allow you

887.4 -> to connect to desperate data sources wherever they may live.

893.25 -> Our motivation behind Data Connectors is

896.14 -> to truly embrace the brownfields

898.2 -> and create a federated interface that acts

902.28 -> as an abstraction over the underlying data sources.

906.4 -> It fuses data from IoT sensors, cameras, documents,

911.94 -> maintenance history, other context setting

913.83 -> and business applications

914.93 -> and it's able to fuse context

916.64 -> between all these different data sources.

919.3 -> And how it does that, it creates a knowledge graph

925.62 -> using these data sources.

928.08 -> The Data Connectors come built in with AWS services.

932.95 -> For example, we have out of the box connectors

935.5 -> for AWS IoT SiteWise,

937.49 -> a service that stores industrial IoT data.

941.81 -> It connects seamlessly with Kinesis video streams

944.87 -> that allows you to ingest and store video data from cameras,

949.87 -> as well as allows for live and historical playback.

953.61 -> And while I'm on the topic of KVS, Kinesis video streams,

957.54 -> we also just announced a Greengrass Edge Connector

961.44 -> for KBS that now allows you to control

965 -> what video from cameras is uploaded to the cloud

967.77 -> for cases where you have bandwidth constraints,

970.13 -> or you want to limit the amount of video

972.09 -> that you're storing in the cloud.

973.33 -> So it enables you to upload data based on certain triggers.

978.17 -> The trigger can be a user requested a video segment,

981.91 -> it can be an alert has happened in the facility.

985.3 -> And because then you can use that as a trigger,

988.67 -> to only upload certain segments that relates

991.64 -> to the timestamp of the event.

993.35 -> So a really cool tool to add video to your digital twins,

997.08 -> but do that in a cost effective manner.

1000 -> And we also have connectors to Amazon S3,

1003.29 -> which we know is a typical place

1006.02 -> where customers are storing application data

1008.63 -> as well as unstructured data, but it doesn't end there.

1012.92 -> We know that there is so many data sources out there

1015.7 -> that customers wanna connect to, and as we go forward,

1018.4 -> we're gonna add support for more data sources,

1020.827 -> but we wanted to make it easy for developers

1023.1 -> to get started creating connectors

1025.66 -> for their own data sources.

1027.5 -> So we added a flexible framework that allows you

1031.24 -> to write your own data connectors to data sources

1035.55 -> that you may be using.

1037.55 -> We have samples available for data sources,

1040.31 -> such as Amazon Timestream, Siemens MindSphere,

1044.3 -> as well as Snowflake,

1045.68 -> and you can easily take these connectors

1048.17 -> using our custom connector framework

1050.81 -> and apply them to any of the data sources

1053.69 -> that you wanna connect to your digital twins.

1060.43 -> Next, we'll talk about the Model Builder.

1063.33 -> This is how you describe relationships and data properties

1066.38 -> for your digital twin.

1069.06 -> Our Model Builder embraces

1071.2 -> the entity component architecture.

1074.06 -> This is any component architecture is commonly used

1078.16 -> in game development,

1079.46 -> but as we talked to customers and they described

1081.59 -> all these use cases to us,

1083.27 -> we realize that this is a very good fit

1085.77 -> for modeling digital twins as well.

1088.97 -> Simplistically, what this architecture allows us to do,

1092.29 -> is model elements in your environment as entities.

1098.02 -> This can be equipment, it can be processes, people, spaces,

1103.24 -> and you attach various components to that entity,

1107.71 -> and that component brings context into that entity,

1111.32 -> so based on the state that each of those components is in,

1116.09 -> it drives the behavior for the entity

1118.25 -> that they are attached to.

1120.02 -> So all the data connectors that I talked about

1123.1 -> in my previous slide are now essentially components

1126.93 -> that you attach to entities,

1128.48 -> and these entities can be pieces of,

1130.76 -> it can be assembly lines,

1131.84 -> it can be processes that you wanna model

1134.68 -> into your digital twin.

1135.99 -> So it's a really flexible way

1137.5 -> to build out your digital twin model into a knowledge graph.

1143.62 -> And this knowledge graph is essentially common metadata

1147.57 -> that allow, that connects the dots

1149.112 -> between all these different pieces.

1153.6 -> What this common metadata or knowledge graph allows us

1157.26 -> to do is provide a single unified API

1162.9 -> on top of our knowledge graph

1165.33 -> that essentially removes the heavy lifting

1169.64 -> from the application development side to how you query

1174.7 -> for data coming from your digital twin.

1177.4 -> So all you have to do is call TwinMaker APIs and ask for,

1181.46 -> I want data for this entity from this time to this time,

1185.08 -> and the complexity of knowing where this data lives,

1188.72 -> what data store it is stored in and how to query

1191.51 -> that particular data store is abstracted away.

1194.56 -> So it's a really powerful tool

1195.91 -> for rapid application development using these simple APIs.

1200.82 -> Another thing that the common metadata allows us

1203 -> to do is put all these data streams on a common timeline.

1207.85 -> That means you truly have the power to rewind time

1211.45 -> and see the state of your environment

1213.41 -> or your facility at that particular point in time.

1216.72 -> And rewinding back in time doesn't just do that

1220.01 -> on a data stream per data stream basis,

1222.48 -> it essentially rewinds time across all your data.

1225.92 -> So you can truly see the state of your environment

1230.88 -> in the past.

1232.84 -> And at the heart of a lot of the stuff

1234.47 -> that we have done with a TwinMaker is

1236.71 -> to make managing change easy

1239.57 -> because we know physical spaces

1241.66 -> and physical environments change all the time.

1244.01 -> So with this Model Builder, as your operations mature,

1249.61 -> and you start to collect different types of data

1252.44 -> or new data streams,

1254.55 -> what that looks like is you're adding additional components

1257.9 -> to those same entities without having to redo the graph.

1261.24 -> So it's really easy to keep your knowledge graphs up to date

1264.87 -> and as things progress, bring those improvements

1267.43 -> into the virtual digital twin as well.

1273.2 -> Get a sip of water.

1282.58 -> Okay.

1284.5 -> Next we have the Scene Composer.

1286.89 -> Scene Composer allows you to create

1289.08 -> those 3D spatial representations of your environments

1293.38 -> with widgets for data.

1296.734 -> We've embraced importing existing models

1301.13 -> into the Scene Composer to create your visualization.

1304.3 -> So we wanted customers to be able

1305.77 -> to reuse their investments in a CAD or 3D modeling

1309.66 -> that they have already invested.

1311.55 -> So you can bring in existing CAD models

1314.51 -> or point cloud scans or other scanning models

1317.94 -> and convert it into the glTF format.

1321.28 -> The reason we're embracing glTF format is,

1324.55 -> one, it's lightweight, it's open-source,

1328.16 -> there is no licensing, so it's just very easy

1331.33 -> for you to get started using it.

1333 -> And it's also an exchange format

1335.49 -> that has been optimized for the web.

1338.37 -> So it's really easy for you to bring your existing models

1342.66 -> of different formats, convert them to glTF,

1345.02 -> optimized for the web and start building your 3D scenes.

1349.96 -> What we have in the Scene Composer,

1351.89 -> it's a very simple 3D composition tool.

1356.17 -> So it doesn't have all the bells and whistles

1358.85 -> of a 3D editor,

1360.05 -> its purpose is not to create 3D models from scratch.

1364.38 -> It is a composition tool that allows you to place

1367.98 -> and position different types of equipment

1370.65 -> into the same scene.

1372.15 -> So you can compose an environment

1374.94 -> that mimics the physical world.

1379.2 -> And once you have that base scene created,

1382.5 -> you can add data overlays on top of that.

1385.36 -> Data overlays are widgets

1387.46 -> that you add on top of your scenes,

1390.13 -> that connect back to the data that you've modeled

1393.27 -> in your knowledge graph.

1394.76 -> So you can add tags that connect back to time series data,

1398.9 -> and you can define simple rules

1401.4 -> that tell you how that the visual

1403.74 -> for those annotations need to change as your data changes.

1407.19 -> So for example, if a temperature exceeds 50 is a rule

1410.51 -> that you've configured,

1411.64 -> you can easily see that it changes in real time

1414.51 -> to depict that this particular equipment

1416.61 -> that this tag has been added to, it needs some attention.

1421.96 -> And again, we've made it easy to manage change,

1424.49 -> and that is possible because of the composition nature

1428.26 -> of these scenes.

1429.15 -> So if you think about a piece of equipment needs

1432.36 -> to move from one room to the other,

1434.16 -> you don't have to create a monolithic model from scratch.

1437.79 -> All you have to do is select that piece

1439.33 -> of equipment and move it.

1440.86 -> And you have a newer version of your scene

1443.35 -> and you can move forward from there.

1448.7 -> And then finally, I'll talk about the Application Toolkit.

1451.74 -> So this is how we want developers to be able

1455.4 -> to create end user applications really fast,

1458.61 -> where we launched a plug-in application plugin for Grafana.

1464.06 -> This works in both open-source, self-managed Grafana,

1468.08 -> as well as Amazon managed Grafana.

1470.33 -> If you wanna not have the heavy lifting

1473.843 -> of managing and spinning up your own instances.

1478.22 -> And it's based on a front end developer SDK

1482.25 -> that we wanna make available for our customers

1484.31 -> and partners to extend these digital twins

1487.41 -> into their own custom applications or for our partners

1490.73 -> into their solutions as well.

1493.16 -> So we want it to be really flexible and open-source

1495.322 -> in that case.

1497.36 -> And we're also embracing open-source a lot

1499.86 -> because there is so much innovation happening

1502.09 -> in the digital twin space.

1503.38 -> So we wanna make sure that there is a place

1506.032 -> where we are constantly adding samples

1510.43 -> and tutorials and instructions of how you can create

1514.05 -> digital twin applications

1515.74 -> that are specific to your use case.

1518.14 -> So we have a GitHub repo that stores all of the sample code

1521.79 -> and tutorials that we have available today.

1524.22 -> And we will be constantly adding more over there

1528.34 -> as well as integrating with AWS partner tools

1531.07 -> and solutions available there as well.

1536.09 -> I wanna do a quick plug for Amazon Managed Grafana.

1538.73 -> It was launched in general availability earlier this year.

1542.32 -> It's a really cool tool.

1544.43 -> Grafana is a really cool tool

1546.37 -> because it gives you the flexibility of data visualization

1549.73 -> across many different sources.

1551.58 -> And with Amazon Managed Grafana, you are focusing,

1555.386 -> you're free to just focus on building those applications

1558.33 -> without having to worry about maintaining

1561.29 -> and creating and spinning up

1562.46 -> and spinning down in those instances.

1565.74 -> Here's another screenshot

1566.92 -> of the cookie factory example built up in Grafana.

1575.26 -> So if all of this is sounding interesting to you

1577.67 -> and you wanna get started, it is a fully managed service.

1581.36 -> So you can log into the AWS management console,

1585.49 -> navigate to the TwinMaker console,

1587.53 -> and you can get started instantly.

1590.53 -> We also have, the GitHub repo that I talked about,

1593.64 -> has a fully built out cookie fact,

1595.38 -> the one you saw available there as a sample.

1599.45 -> So you can download that and get started

1601.68 -> with your own cookie factory in your own account.

1605.775 -> So you can use that to explore the various features

1609.02 -> of the service, but the team's done an excellent job

1611.6 -> to make it opinionated, but extensible.

1614.22 -> So you can easily replace the datasets

1617.23 -> in there in the cookie factory with their own data sets

1619.56 -> to create your digital twins.

1622.47 -> And the GitHub repo, as I mentioned,

1624.1 -> has tons more samples available.

1626.65 -> It has samples

1627.483 -> if you wanna connect machine learning inferences

1630.69 -> from SageMaker for your data streams

1632.91 -> that have been modeled in TwinMaker,

1635.09 -> the sample manages the transport of the raw data

1638.87 -> to the inference end point

1640.3 -> and then transport the inference result

1642.4 -> back into your data stores,

1644.06 -> bringing that intelligence into your digital twin.

1647.94 -> Same thing with, if you wanna run simulation based insights

1653.27 -> from our partner tools, also samples are available there,

1656.95 -> and you can run these ML models

1659.26 -> and simulation models in real time.

1661.92 -> So using streaming applications,

1664.67 -> and those are all available

1666.47 -> for you to try out and contribute.

1669.94 -> You know, if you're making something really interesting,

1672.46 -> bring it back into the community and we can all share

1676.2 -> all the good work we're doing here.

1678.944 -> And I wanna call out our partners.

1683.58 -> We understand there is so much innovation happening

1686.52 -> in the digital twin space.

1688.13 -> Our partners are doing an excellent job

1691.16 -> at the bleeding edge, bringing all this innovation

1693.85 -> to our customers.

1694.99 -> So we're really excited about our launch partners,

1698.54 -> you can see them over here.

1702.27 -> We have a kiosk available at the Venetian sands expo.

1707.24 -> It's 1957, so if you wanna talk to our partners,

1710.78 -> you wanna see how partner, oops,

1714.16 -> that went faster than it needed to.

1719.24 -> So yeah, if you wanna see these partner integrations

1722.49 -> in action, please come stop by our booth.

1726.39 -> Myself and my team are going to be

1728.99 -> at the booth 1957 again, tomorrow morning at 10.

1733.03 -> So if you're around, please come say hi,

1735.82 -> and look at all the demonstrations

1738.13 -> that our partners have created.

1741.02 -> I wanna call out.

1742.24 -> There is so much diversity in the solutions

1745.9 -> that our partners are bringing to the table.

1747.87 -> I don't know why this keeps happening, sorry.

1752.86 -> Yes, very, very excited about our partnership

1755.54 -> with Siemens to bring their

1757.98 -> comprehensive digital twin portfolio.

1760.21 -> Along with TwinMaker partners like Embassy Of Things

1763.93 -> and Element Analytics bringing together the IT and OT world

1768.31 -> for your digital twins.

1770.41 -> Consulting partners that are ready

1772.93 -> and trained to work with you on your specific use cases

1777.74 -> and customized digital twin solutions.

1779.77 -> They have the domain expertise

1781.36 -> and can provide accelerated deployments.

1786.28 -> as a teaser, I wanna show you a quick video

1789.66 -> of the work we're doing with Siemens.

1792.69 -> This is a Cold Brew Coffee Plant,

1796.16 -> and we're really excited about this

1798.81 -> because with this partnership,

1801.27 -> you can take the capabilities

1802.91 -> that I talked about with TwinMaker

1805.13 -> and marry that with the expertise

1809.24 -> that Siemens has across this portfolio

1811.87 -> of digital twin capabilities and bring them together

1815.07 -> into a single pane of glass.

1817.07 -> What you're seeing here is a Mendix application.

1820.22 -> Mendix is a tool that allows you

1822.02 -> to rapidly build applications.

1823.92 -> And it shows the capabilities from TwinMaker,

1827.73 -> which the scene viewer, all the data that's connected

1831.63 -> to the knowledge graph.

1832.53 -> In this case, the data is coming from MindSphere

1835.95 -> as well as interactions

1837.55 -> that enable you to immediately switch to a detailed context

1843.61 -> about these assets

1844.79 -> or this coffee brew line coming from tools

1847.83 -> such as Simcenter, or coming from PLM tools,

1850.9 -> such as Teamcenter.

1852.35 -> So you are able to really bridge the gap

1854.81 -> between going from a facility view

1857.63 -> into a deep understanding of the piece of equipment

1861.25 -> that you are trying to diagnose an issue

1864.19 -> with or whatever the use case is.

1866.29 -> So we're really excited about this.

1867.83 -> If you wanna learn more, again,

1869.2 -> please come visit the kiosk, it's 1957.

1873.32 -> And with that, I will hand it off to Dane.

1878.1 -> Welcome Dane. (audience applauding)

1882.165 -> Thank you, Saras,

1884.897 -> Can you guys hear me?

1886.06 -> Awesome.

1887.08 -> Cool, so my name is Dan Loughlin.

1888.41 -> I'm an innovation engineer with Invista.

1890.13 -> Invista's a polymer and intermediate company,

1893.13 -> and we make a lot of the materials

1894.57 -> that ended up in products that we use on a daily basis

1897.61 -> and have come to know and love.

1899.68 -> So instead of boring you guys with a bunch of talk

1903.26 -> about chemical processes and all that fun stuff,

1906.14 -> I'm gonna play a fun marketing video for you,

1908.7 -> and it'll overview Invista.

1909.533 -> And then we can talk about how we intend

1911.77 -> to use digital twins.

1914.93 -> Oh, cool.

1920.2 -> At Invista,

1921.66 -> continuous improvement drives our people

1923.66 -> toward constant innovation,

1925.65 -> working together around the world

1927.44 -> to solve life's challenges, just people helping people.

1932.4 -> Not simply to take groundbreaking products to market

1935.52 -> or license advanced technologies for the benefit of others,

1939.43 -> but to make family time more fulfilling,

1942.35 -> free time more exciting, and bedtime a little more cozy.

1948.11 -> We're in your cars, your carpets,

1951.35 -> even the fiber in your clothes,

1953.54 -> because we are in the ingredients that make it all possible

1957.34 -> and so much more.

1959.31 -> We're in a lot of the things you care about

1962.48 -> and in a lot of the things that take care of you,

1967.06 -> but ultimately we are invested in making the world

1970.24 -> around you a better place

1973.38 -> to work in,

1975.12 -> to live in and to thrive in.

1979.31 -> Are you in?

1985.66 -> Awesome.

1986.84 -> Well, so now that you guys kinda know a little bit

1988.63 -> about Invista and have an introduction to us,

1991.11 -> I wanna talk about why digital twins are important to us

1993.94 -> and how they help achieve the vision that you just saw.

1997.22 -> So as a manufacturing company,

2000.17 -> we have a lot of people who are doing

2001.66 -> a lot of different jobs and have different skill sets.

2004.32 -> And so there's a couple of challenges that we face

2007.07 -> when we want to connect the workers who are out

2010.55 -> in the field to the data that can actually enable them

2013.06 -> to make decisions faster.

2014.99 -> And one of those is that our assets

2017.49 -> or our data assets are in siloes, right?

2019.84 -> And so much like many of you guys probably there's

2022.95 -> four or five different platforms that people might use,

2025.77 -> ones for accessing work order instructions

2028.41 -> and others for accessing IoT data.

2030.94 -> And other might be for accessing a 3D model

2033.67 -> or piping and instrumentation diagrams,

2038.45 -> all these different things.

2039.84 -> And so people to Saras' point earlier,

2042.45 -> are context switching back and forth.

2044.07 -> So they're trying to understand if I'm looking

2046.17 -> at this on this diagram,

2047.66 -> and then I look at the data is what I'm seeing

2050.27 -> in the diagram coming from the same place

2052.17 -> that my data is at that I'm looking at

2054.57 -> in this other database.

2056.12 -> So we have a real challenge there.

2059.4 -> And to the kind of resolution point, the goal,

2062.298 -> especially in chemical manufacturing want people to be able

2065.26 -> to make really good decisions the first time.

2068.59 -> A lot of times people have to make critical decisions

2072.47 -> and we wanna make sure they have the best data

2074.43 -> to enable that decision-making process.

2078.16 -> And then from the context of an overall plant,

2083.44 -> a lot of times, we're limited to whatever the job role is.

2086.322 -> I'm gonna take this off.

2087.98 -> We're limited to whatever the job role is that we're focused

2091.15 -> on at the time that we're working on it.

2093.63 -> And so, for instance,

2095.61 -> if I need to get an understanding of kind of cohesive image

2099 -> of an entire facility, it's really hard for me to do that

2102.66 -> if I'm looking at a dashboard

2103.91 -> that only covers motors, right?

2105.82 -> And so the context is really important because,

2110.08 -> particularly in process type manufacturing facilities,

2114.87 -> everything's interlinked, right?

2116.52 -> So if there's an issue down here,

2117.8 -> more than likely you might have an issue downstream.

2120.25 -> And so all these things affect each other.

2121.77 -> So seeing this kind of cohesive whole

2123.88 -> and getting an understanding of that is really important.

2127.84 -> So I'm gonna kinda walk through an example

2131 -> of where somebody would use something like TwinMaker.

2135.74 -> So let's say I'm an operator out in the field,

2138.87 -> I'm doing my routine walkthrough for the day,

2141.43 -> and you might see like an oil fil that's low.

2146.69 -> The challenge is I don't know

2147.76 -> when the last time the oil was changed.

2149.82 -> I don't know if there's a crack

2151.09 -> or if it's just that it needs maintenance.

2153.96 -> There's a lot of things that I don't know

2155.15 -> about the situation that's going on.

2157.38 -> And if you use the kinda analog version

2159.76 -> or what we've done traditionally

2161.51 -> in the past is we've had to go back to our desk,

2164.86 -> sit down, open the software that we use,

2168.54 -> find the piece of equipment that I'm looking for,

2170.44 -> type it all up, submit it, go back out in the field,

2173.03 -> find more information, potentially fix it or not.

2176.13 -> So there's a lot of steps involved there

2178.81 -> and what the digital twin stuff allows us

2180.96 -> to do is we can now essentially serve that up

2183.99 -> to the person at the time that they are looking

2186.15 -> at the asset that they're gonna be working on.

2188.01 -> And they can actually submit data from the field as well.

2190.56 -> So the new solution to this would be, I walk out,

2194.81 -> I see the oil level's low,

2196.97 -> from my phone I pull up Grafana or TwinMaker,

2201.74 -> I put in a SAP work order that says,

2203.927 -> "Hey, this is what I need."

2205.71 -> I check the work order history and then say,

2207.697 -> "Okay, I recognize that it's probably not a leak

2210.28 -> because the oil hasn't been changed in a while."

2213.42 -> And then everything that I need is right there

2215.68 -> in front of me.

2216.513 -> I don't have to walk back to the desk, I don't have to go

2218.1 -> and ask somebody for SAP access,

2219.84 -> I don't have to do any of that stuff.

2221.5 -> And so it's really valuable to have this kind

2224.15 -> of abstraction layer where you can aggregate a lot

2227.107 -> of the data together in one piece.

2230.6 -> And then the other thing that we're trying to address

2232.68 -> with the digital twins is,

2235.9 -> and I'm sure you guys may be experiencing this as well.

2239.2 -> You know, we're, we're losing talent,

2240.7 -> both in the form of retirement.

2242.25 -> And these folks are masters of their craft, right?

2245.81 -> They have a lot of experience.

2248 -> And so how do you institutionalize that experience

2250.25 -> so that people can learn from that and be as effective?

2254.98 -> And I have a funny story that I love telling around this.

2259.45 -> One time I went to a facility

2261.51 -> and there was a guy and I won't name his name,

2264.62 -> but he had been there for like 40 years,

2267.97 -> obviously was one of the most senior people there,

2270.84 -> knew the plant like the back of his hand.

2273.1 -> And so I went on a round with him and I'm following him

2277.47 -> around and we stopped in this area and he kinda stops

2280.68 -> and like moves his head.

2283.99 -> And I'm like, "What's going on here?"

2285.16 -> You know, we're standing in the middle of a chemical plant.

2287.37 -> And he says, "Well, you know, something sounds off."

2289.82 -> I'm like, "We have earplugs in, man.

2291.67 -> Like, I don't know what you're talking about."

2294.1 -> And he goes, "No something sounds off."

2295.88 -> I was like, "Okay, you know, so what's going on?"

2300.73 -> And he walks over and he's feeling the pumps

2306.26 -> and one's apparently vibrating differently.

2308.95 -> I can't tell the difference, but, you know,

2310.51 -> he knew that one was vibrating differently

2312.8 -> and he pulls a screwdriver out of his pocket,

2316.2 -> sticks it on the motor and puts his ear to it.

2319.04 -> And he goes, "Yup, race bearing."

2322.36 -> And I'm like, "Okay, clearly you know something

2325.91 -> that I don't, right?"

2327.77 -> And so that's a good example of,

2330.74 -> these are experiences that, I mean,

2332.787 -> he's the original connected worker, right?

2336.28 -> His connection just happens to be that he spent 40 years

2339.1 -> and knows it really well.

2341.13 -> So what we're going,

2342.3 -> or what we're trying for is how do you create a tool

2344.5 -> that augments people's experience

2347.28 -> so that they can kind of understand the same context

2349.95 -> as what that operator did when they put the screwdriver

2353.73 -> to a motor to hear for a race bearing,

2356.44 -> how do you actually provide that context through data

2358.837 -> and a visualization so that it doesn't take 40 years

2361.74 -> for people to be able to understand what they're looking at?

2367.13 -> So, as we kind of recognize, we said,

2370.787 -> "Hey, you know, we know that we need to address this.

2375.267 -> And we know that we need tools

2376.6 -> that help us do this more effectively."

2378.45 -> We started scoping out if we wanted to build something

2380.54 -> like this ourselves, what'll this take?

2382.54 -> And we recognized that as a manufacturing facility,

2385.54 -> we don't have the technical resources

2388.31 -> to throw 50 people starting up a platform

2390.91 -> to be able to support this across the business.

2394.93 -> So we were thrilled when Saras

2396.367 -> and the team at AWS reached out to us and said,

2398.627 -> "Hey, you know, our goal is

2400.08 -> to create this aggregation layer

2402.49 -> where we can actually overlay data on top of a 3D model."

2405.78 -> And so they've been a great partner

2407.13 -> for us working through,

2408.92 -> how does this actually get implemented

2410.39 -> in an industrial company?

2414.36 -> Sorry, so effectively what we worked with them

2418.5 -> on are these three things.

2419.76 -> So we need to connect the dots between data sets.

2423.16 -> As Saras mentioned, it's important too that we don't have

2425.76 -> to aggregate 'em all into one database.

2427.64 -> Like we wanna be able to access the data

2429.33 -> where it lies because we have it all over the place, right?

2432.66 -> And who knows what kind of relationships can be drawn

2435 -> between different pieces of data?

2436.68 -> And so having that be accessible is really important.

2440.27 -> We wanna build a contextualization platform.

2444.38 -> So we want 3D models.

2445.72 -> We want people to be able to look at the models

2448.317 -> and understand, okay, when there's a race bearing going out

2453.6 -> and something's flashing,

2455.07 -> I can look at the motor that it's actually going out on,

2457.72 -> and then I can understand, okay,

2459.67 -> based on where it's at, I might need a tool cart.

2462.15 -> I might need, you know, X, Y, Z tool, et cetera.

2465.6 -> And so there's a lot of context that can be provided

2467.61 -> by having a 3D asset in front of you to look at.

2471.33 -> And then the last thing that's really important

2473.71 -> to us as a company is this concept of citizen developers.

2478.31 -> So what we recognize is that if every dashboard

2483.11 -> and digital twin relies on Dane to set up for people,

2486.92 -> we're gonna be here for the next 20 years waiting

2488.93 -> on Dane to be done, right?

2491.16 -> And so what we needed is we needed something

2493.62 -> that's easy enough and accessible

2495.76 -> for people who are interested in the space

2499.13 -> to jump in and be able to make the interactions

2503.2 -> or be capable of creating the partnerships

2506.61 -> or relationships between the model and the data

2509.42 -> that we're talking about.

2511.02 -> So a big key thing here or a key message here is not

2516.22 -> that we're looking to displace people with digital twins.

2519.86 -> What we recognize is that digital twins allow us

2522.61 -> to essentially focus people on what they do best

2527.09 -> and get rid of a lot of the kind of context switching,

2531.33 -> cognitive friction, all these different things

2533.53 -> that people deal with on a day-to-day basis.

2538.46 -> So here's an example of our actual facility in TwinMaker,

2542.513 -> and there's kinda four key pillars if you will,

2546.296 -> that attach here.

2547.697 -> And one caveat here too,

2550.07 -> you'll see when we jump in,

2553.08 -> there's a kind of like a live view,

2556.6 -> or I shouldn't say live 'cause it's not live,

2558.61 -> but it's a high definition perspective.

2561.84 -> And then this is actually the CAD model

2564.652 -> or a GLB model, if you will.

2567.43 -> And so the thing that enables this is

2569.36 -> from an asset information perspective,

2571.03 -> we use OSI PI.

2572.74 -> OSI PI allows us to aggregate data

2574.79 -> from a lot of IoT sensors into a place

2577.53 -> where we can put it on dashboards

2579.67 -> and actually deliver it to people.

2581.58 -> And so we're pulling from OSI PI here.

2583.61 -> You can see on the right hand side,

2584.83 -> there's live data coming in from the field.

2588.38 -> And then we're using Snowflake

2589.94 -> as well as essentially a data warehouse

2593.14 -> to make it easy for transmission and that kinda thing.

2597.4 -> And then you have TwinMaker who comes in

2599.83 -> and they allow us to do alerts.

2602.01 -> So we can use DCS alerts,

2604.81 -> or you can use a control room.

2607.23 -> You can customize your own

2608.28 -> and you can build rules around this.

2610.28 -> So in this situation where it turned red,

2613.53 -> it was because it was out of spec for a rule

2615.59 -> that I set for that piece of equipment in my dashboard.

2619.77 -> And then the final piece of this is Matterport.

2624.12 -> One of the biggest challenges

2625.51 -> in the digital twin space is being able

2628.97 -> to create 3D models that are actually portable

2631.87 -> and accessible for people.

2634.31 -> So if you think about point cloud,

2638.22 -> that was one of the first places we started

2640.07 -> as we generated all these point clouds.

2641.63 -> And it's really cool

2642.463 -> 'cause you got like a trillion dots, right?

2644.45 -> And it's real exciting.

2646.1 -> The problem with that is it might be a terabyte in size

2650.18 -> and nobody's gonna download that on their iPhone

2652.17 -> and use it in the field, right?

2654.21 -> And so we needed a lightweight solution

2655.93 -> that allowed us to give people the context that they need,

2659.35 -> but also not destroy the bandwidth of the site

2662.89 -> because we're downloading models all the time.

2665.73 -> And so this is where Matterport comes in

2667.51 -> and they've been also a great partner for us.

2669.48 -> And we're really excited

2670.33 -> about the collaboration that's happening here.

2673.02 -> So for you guys that may not know,

2674.663 -> this is a video of Matterport

2677.64 -> and all the different kind of modalities you have.

2680 -> And so this is the kind of pretty perspective

2682.98 -> as you can see, this as our actual facility,

2685.68 -> it's in high resolution and what we've done

2688.56 -> and what we've used Matterport for

2690.69 -> to this point is we can label things.

2692.44 -> So when you're teaching people about a piece of equipment,

2696.06 -> you can actually show 'em what you're looking for.

2698.45 -> You can create links and do that kinda stuff.

2701.49 -> The thing that TwinMaker kind of enables over the top

2704.5 -> of this is that you also have a data layer, right?

2706.74 -> So you have the context, as far as descriptors,

2710.91 -> you can add links and then now I can pull in IoT, data,

2713.76 -> SAP information, all these things.

2715.68 -> So this is kind of a rough version and I'm calling it out

2718.81 -> because I know it's rough, but the end product will

2722.18 -> actually be the view that you saw

2724.77 -> of Matterport that's high definition,

2728.02 -> completely integrated over the top.

2729.42 -> So we're super excited about that because what that means is

2731.97 -> that everybody that's out on the shop floor will have access

2735.72 -> to high quality models.

2737.35 -> And so, I mean, you can count the, you know,

2741.2 -> or you can read tags and stuff on the labels,

2744.7 -> which is really, like I said, high fidelity,

2747.62 -> and it actually will load at a reasonable pace,

2750.34 -> it's not super burdensome.

2752.28 -> And so the integration between TwinMaker

2754.79 -> and Matterport is huge for us because it allows

2758.43 -> for that kind of synergy between the two platforms.

2766.88 -> So the business outcomes that we're looking for,

2769.71 -> like I said, knowing the context

2772.06 -> or knowing the location of the different assets

2774.12 -> that you have around you improves productivity

2777.38 -> for field operators.

2778.28 -> So, if they get a work order, then it comes

2781.3 -> through on their phone on their email

2782.85 -> and they have a link that says,

2784.9 -> you have a work order that's here.

2786.59 -> I click on it, it takes me to the location

2788.75 -> of the piece of equipment

2790.27 -> that I'm supposed to be doing work.

2791.49 -> I zoom out and I go, okay, it's that motor,

2793.99 -> it's not one of the other 12 motors

2795.78 -> that are in that bay, right?

2797.18 -> So that context provided me the understanding

2799.41 -> of where I'm at, where I need to go

2801.097 -> and all these different things.

2802.74 -> And so that's the context that we're talking about.

2805.73 -> And one of the things that we're most excited

2807.55 -> about is we think that, that really enables people,

2811.5 -> especially if you're walking around

2816.09 -> and have to cover a large area,

2818.12 -> it allows you to kinda really get a good understanding

2820.26 -> of your environment.

2822.37 -> We wanna continue to build excellence

2824.631 -> in environmental health and safety,

2827.06 -> wanna make sure that people are safe.

2828.664 -> And if we can provide additional contextualize information

2832.55 -> that help them make good decisions when they're out

2834.52 -> in the field, that's a really valuable thing for us.

2838.77 -> And we're really proud of our EHS record.

2842.58 -> And we wanna make sure

2843.7 -> that we're leveraging human expertise

2846.91 -> to create value for our employees.

2849.14 -> So we think that there's a lot,

2850.991 -> there are tasks that are burdensome for people,

2854.93 -> and we'd much prefer to offload that

2856.9 -> to a platform and allow them again,

2859.2 -> to do what they do best and do what they're passionate about

2861.48 -> because passionate people do really impressive things.

2868.087 -> And so what's next for us?

2871.12 -> Well, the first one is scale, right?

2873.62 -> So again, with Matterport and others,

2875.418 -> we can ship cameras to facilities

2878.94 -> that have never had 3D models before.

2881.07 -> And people Bluetooth connect to their device,

2883.82 -> walk around and scan the facility, upload it.

2886.27 -> And then we make the connection with TwinMaker.

2888.68 -> So that's our goal is to continue to scale out.

2891.84 -> You know, we have facilities that are scanned.

2894.93 -> We have facilities that are scheduled to be scanned.

2897.28 -> And so that's kinda the next step.

2899.65 -> We wanna foster that citizen developer capability.

2903.09 -> So we want people to look at it and imagine

2905.91 -> what their lives could be

2906.92 -> and how things could be transformed

2908.56 -> if they only could use a 3D model,

2912.03 -> and then we wanna help 'em build it.

2913.84 -> And then we wanna empower them to build it

2915.88 -> for themselves and to share with others.

2918.68 -> What we recognize in this space is that any group

2920.84 -> that you talk to,

2921.71 -> so if you talk to an operator

2923.87 -> versus a maintenance technician,

2926.31 -> versus somebody in supply chain,

2927.82 -> they're all gonna give you different answers

2929.47 -> as to what digital twins are good for, which is great,

2933.38 -> because that means that we have a lot of opportunity.

2936.18 -> It means that there's a lot of value to be had.

2940.84 -> The challenge is that I can't build it for everybody, right?

2944.74 -> And I don't know everybody's job to the degree

2946.79 -> that I need to in order to effectively build that.

2949.66 -> And so that's the next capability

2951.99 -> that we wanna build is, hey,

2953.664 -> how do we help enable you

2955.83 -> to build something that's valuable?

2957.72 -> And then can we share that with people

2959.04 -> in similar type of roles?

2961.53 -> Finally, we wanna make sure that we continue

2963.39 -> to collaborate with the AWS team.

2965.35 -> Like I said, the AWS team's been great

2967.27 -> as far as kinda listening to feedback

2969.11 -> from us and implementing it.

2971.83 -> And we hope that, that collaboration continues

2975.44 -> throughout the industry too,

2976.45 -> not just between the two companies here.

2980.51 -> So with that said, I do wanna say thank you

2984.27 -> for you guys for being here and listening to us.

2987.663 -> I wanna say thank you to the AWS team again,

2990.48 -> and also thank you to the internal Invista team

2992.85 -> that made this possible, I really appreciate it.

2995.61 -> At this point I'll hand it back over to Saras

2997.69 -> and she can finish it out for us.

3000.695 -> (audience applauding)

3006.93 -> Thank you, Dane.

3008.09 -> And thank you everybody for coming.

3009.8 -> I hope this was entertaining, helpful, not boring, at least.

3015.31 -> So, as I said, visit us at the booth.

3018.27 -> I wanna plug a ChalkTalk that's happening on Friday.

3022.26 -> It's IoT 317.

3025.73 -> If you're interested in learning

3027.23 -> and getting a more hands-on information

3029.71 -> on the technical details of the service

3031.77 -> and how you go about building digital twins,

3034.04 -> please be sure to check that out.

3035.42 -> Again, it's IoT 317,

3038.18 -> and please remember to fill out the session survey.

3043.08 -> Thank you all for coming.

3044.37 -> We'll be here.

3045.8 -> We have 10 minutes for this room,

3047.98 -> so we'll be here in case anybody wants to come up

3050.35 -> and ask us questions, I'm happy to chat,

3053.08 -> otherwise, thanks for coming,

3054.78 -> and I hope you have a good evening.

Source: https://www.youtube.com/watch?v=fdBKRyhC9Yk