Virtual Collaboration with VRED Server

LUKAS FAETH: So my name is Lukas I'm the product manager for VRED, as well as VRED Server, which I'm going to talk about today. And as I said, I'm going to talk about virtual collaboration. For those who are here to see something, but the [INAUDIBLE] or something that's around-- I have to explain in the beginning that it's not around that. Although, it might be indicated by the title.

So I've broken down very easily or tried to structure the presentation quite easily. I will talk about 10 minutes on PowerPoint slides after that, roughly 30 to 40 minutes life demo. It will be a Livestream demo, which is connected to some technology challenges, I'd say, especially on an exhibit like that. But we will see. We will see the final result. Maybe the internet gets better over time. So I will just talk a bit. It

So as you can see, quite simple, I will talk about virtual collaboration-- what, why, and how. So what is VRED Server? Why are we doing that? And how are we doing? So the how is going to be the live demo.

So what, what is VRED Server? I think I've introduced that class as a beginning class. That's why I will start from scratch, as I said 10 minutes. For those who know VRED Server already-- this was presented some years ago already in different contexts-- might be a bit bored in the first five minutes. But I promise there's new stuff coming in the live demo.

So what's VRED Server, and what's VRED Server Manager? What does our positioning say? So what is it for, what is it helping with? "It's to speed up informed decision making and manage your hardware resources effectively from one interface. Embrace increasing complexity and pace by providing access to the latest design data at anytime from any place to all stakeholders and contributors of the design process."

So if we break that down a bit, you see we want to speed up informed decision making. It's about managing hardware. It's about being effective, about accessibility. And I'll do my best to break that down in, let's say, the technical context in the next one.

So starting with the Tech Base. VRED Server it's a Data Center based system that's leveraging the VRED engine. So what that means is basically it's using the same render engine. So it's based on all render notes, which you could also use with your cluster. On top of that, there is VRED Server managing that whole thing.

Obviously, with that, you're able to use the data sets. So you're able to use VBP files that people in your company have set up in VRED Pro or VRED Design. So you don't need to have any additional work to get stuff on VRED Server. You get all the configuration options that you know from your files. I will show you in a second what that means.

It's nearly linear scalable. So that's due to the underlying cluster technology of VRED. And it runs on any device. And as I said, that's why it's coming via the network. And we will see. So I got my iPad. If the network is friendly, I'm going to plan to invite some of you to join me with your cell phones. But we'll talk about that later on.

So VRED Server is consisting of two parts. So if Autodesk is referring to VRED server, we tried to keep it simple, because actually it's two parts. Now one is the base technology, which is a Linux based load balancer. That's the cluster head as I just described. And the other one is what you're going to see in the second is the interface for it, so visual web based platform that lets you control set up and worked with this VRED Server.

So the usual workflow around that would be you have your virtual garage, which is basically just a fancy name for storage for your VPB files or prepared cars. So those cars would be loaded up via VSM, which is VRED Server Manager, VRED Server.

VRED Server's distributing this to Render Notes. Those are calculating the deliverables-- In this case, it's the stream-- and piping it back to VRED Server Manager. And then we will see that in the back end interface that I'm going to show you later on.

So why are we doing design collaboration? For those who know VRED Server and the history about it, it's a very open and a very big thing. For design collaboration, we really focused on just one application area to make it easily accessible for everybody, because we experienced that it was so flexible that sometimes people were struggling with nailing down their use case. And that's why we decided to provide a very easy one, design collaboration.

So why would did we do that? Because customers obviously approached us, and they said, we have a need. And the need is to connect people and data. So the overaching need, I think everybody working with OEMs or in the automotive industry is aware that almost all of them have a satellite design studio, for example, at some nice place to live, for example, Southern California or something. And obviously, the headquarters are somewhere. So they are looking to connect the people there and have the discussion more easily than at the moment with VRED.

Whom did we do that for? So it says design collaboration. So first of all, we tailored it for the design center, so really the people working in the design center. And I know that they are way more, all the modelers and everybody. But we tried to keep it simple for that presentation and kept the most important personas.

So on left hand side, you see a designer, obviously. He's about conception. What he would do with that tool is design discussions, for example, with a colleague. So if he did something, prepared it in VRED, loads it up to VRED Server, and then you could use the tool to just communicate about what you're seeing there-- also for design reviews.

I think there are bigger design reviews which are happening on ray tracing clusters, on Powerwall. This is not what we are targeting here. So this is not replacing your Powerball presentation, because I think that is more for ad hoc discussions between people working in that department.

Then we have engineers, obviously, taking care of a realization. So not all the time when a designer comes up with a fancy idea, it's realizable. So that's what those guys are taking care of. And we've got a manager who is more or less like the department lead. And then there is a decision maker, the people that come in and just get a presentation, decide something. And based on that decision, their work will be applied to the current project.

So we have very different set of people here. I think we can agree that the two on the left side, with a bit of practice, they would be able to use VRED without an issue. They hopefully would have the right hardware to do it, maybe the designers not.

But so it's more realistic to use VRED Pro for the two personas on the left, while I think for the two personas on the right, there's not even the intention to be a VRED specialist. So that is why we try to make it easy. So all of those four people or personas could communicate without having the issue of accessibility hardware and all that and the required skill set obviously.

So the challenges we saw in the process was time, as always-- so in this case very specifically, the time to prepare a session, for example, or how to organize that, as well as to transfer the data. So for those who know VRED Pro, we have a feature that's called Connector, which is connecting two VREDs. But you would have to have the same VBP file to judge the same stuff.

So VRED will not tell you that the other person is looking at another file. It's just syncing the camera and some functions in VRED. So that's why when you had to transfer the data to a colleague. That is something that was challenging that we tried to solve for that one. Also connectivity, I think that's the key need that only to say too much to that. That was the overarching need we were discussing a few minutes ago.

And it needs to be process integrated. So the results need to be fed back into the process or the pipeline. So somebody, an operator for example, could use what comes out of that session and works on it. And compatibility issues or challenges-- so for example, having a different version or having different hardware.

And the requirements, we took a look at the personas, at the needs, at the challenges. And then we came up with the requirements we need for that tool. So it needs to be easy, I think that was clear from the personas already. It needs to be easy to set up and administer so also the IT person or whoever's taking care of working with the system is able to do that.

So we tried to keep that as simple as possible as well. And that is the biggest point for this feature or for this product, it's easy accessible. So it's not requiring heavy hardware and all that.

Connected to that ubiquitous device agnostic so you could run it on an iPad or on a cell phone or on your laptop or on your workstation, wherever you like. Obviously, our customers wouldn't allow a step back in terms of quality. So if we came up with, OK, we will render on the iPad or whatever, they would be like, OK, no, that's not acceptable. So they want to have the same quality as in VRED.

With streaming-based technology, we tried to get the same fidelity, which we usually do. As I said, I think everybody is aware of compression if the internet is too low. So we see stuff like that later, that's happening because of the internet. If you have a stable and fast connection, this shouldn't be a problem. You will have exactly the same result as you know it from VRED Professional.

And flexibility for our customers, especially cloud, especially in design is not a very trustful topic at the moment. That's why we always leave the flexibility. You could do it on Premise. You could do it in the cloud.

If you're ready to go to the cloud, we'll be ready to run it. So the demo will be running on AWS. So what I'm showing is completely cloud today, but you can also do it on Premise. And connected to that is, obviously, a very high standard of security, as well as a process integration that's needed.

So coming back to the last slide of my PowerPoint, I think this is breaking down the positioning statement we had at the beginning. So the main point is to share designs quickly and easily. It's pointing out that you don't need an install, like a software. You could invite your stakeholders to that. And what it is all about is to accelerate informed decision making by accessing the data from everywhere at any time.

So that was the boring part. And the how will be shown in the VRED Server Manager. So as you can see just quickly, we are running on AWS, not very much hardware. So we've got one note for the VSM and we got two GPU notes.

I'm going to show you GPU streaming. You could also do CPU with streaming. You could also cluster CPU stream, for example. So all of that is possible.

It's very lightweight setup, because the data set that I've chosen is also not a very big one. But I think in the context of design, it's not always like the finished, fully-fledged data set. And that's why I think it's quite realistic scenario. So let's firstly just go through the interface.

And you can see, the web page-- just to show that first-- is fully responsive. So if we scale that, you can see it's adapting, obviously. So that would be the layout for the iPhone, for example, for any device that's too narrow to show the wide screen, but adapting to whatever device you are showing it on.

So on the left side, as you already saw, there is a menu. This looks quite complicated at the moment. That is because I'm an administrator. So I would be the IT person for now. So all that monitoring stuff is something that a user don't need to bother with.

So what we can do, just to give you a brief overview, you could create a user profile, obviously add some details there. So you can fold in and out the menu, depending on what you like. There is a dashboard, and I think that's the most important thing for people using design collaboration. As I said, the IT guys will have to cope with other stuff. But if you're just about running collaboration sessions or setting up collaboration sessions, I think this is where you will be mostly.

To quickly talk about the structure, I don't want to bother you too much with that, because this is also technical stuff. I try to keep that as simple as possible But in terms of infrastructure for the setup, we have, in this case, as you can see up here, there's something that says production.

So this is a hierarchical structure that we have in VRED where you could make a very high level split so for example. So for example, if you really want to protect some stuff and want to have divide two departments, for example, you could use that. So nobody of one department could access the data of the other department by accident. This comes originally from configurators.

So what we try to achieve here was to have one that is outward facing, one environment that would be like the live environment that your customers could access while you have a draft or experimental environment and to not accidentally publish a file that you don't want to publish. We have that very high level split so that cannot happen.

Then you need a VRED server. And in this case, we just have one VRED server. Depending on how big or how many concurrent users we want to have, there might be at the point where you need to have two. But we're able to run quite a lot of notes on one on one frets over at the moment.

And just quick, as I said, I just briefly want to discuss those repositories to let you know what they are. I will not go into detail. So we got note repositories. We got image repositories. And we have seen repositories. So basically this is just a place to register your notes, to save your images, and to store your scenes. That's basically it.

This needs to be set up one time when you're setting up your VRED Server. And after that, you don't need to cope with that anymore. So below that's VRED Server, you have note groups, which are the boxes here. And you've got an overview.

So later on if we are running a collaboration session, you will see that the file will be loaded onto those note groups. And you will see a preview. And you will see a status of the note. So this is the note group.

In this case, it's easy because we just have one note. So we don't even need to have a note group. But you can see that there is, the IP, the status. So it's going to be offline, but it's ready to be used. And I think that's the important thing. So it's just waiting for a collaboration session to be started.

And on the other hand I have a second note group, which is done for preview. And I'll go to show you in a second what that means. But this shows you already that you need to set up the notes in a way that you anticipate your internal customers, colleagues, whatever, to use it.

Just to provide an overview and to close the point on the Dashboard, obviously there is an overview about the collaboration sessions at the lowest part. But in this case, I just have one setup. So you can see this is a collaboration session for me, also set up for me. It has a title and a preview with which file we are going to discuss or what the topic might be.

So this is scheduled for November 17th starting at 1:00 PM. The current status is idle. So it's waiting to be started, or it will automatically when the time is reached. And it's currently offline. I could also start it manually. I could join the session, I could alter the session, or I could delete it. So that's basically it for the Dashboard overview.

Let's just go to the to the scenes. So you can see, not very fancy. Two of them are the genesis that we're shipping with VRED, the standard data set. One is the evo. As I said, this is just for demo purposes. So this is where you have an overview about your scenes. Obviously, as I said, this is a very reduced example. That would be way more scenes of you using that productively.

We got collaboration sessions. And there you can see that I've tried to prepare that in a way that it kind of looks like it was used already. So we did some test sessions. As you can see, they are based on the same data set or on different ones. And they are adding up.

So the information is stored throughout the session. Or if you're using a continuous session, you will also keep the information that you're storing there.

Just a quick one, it would be also possible to schedule render jobs by efforts of a manager. This was done as on customer wish some years ago. So this is quite reduced. It's just reading sequences, just if you're interested in that.

To give you an overview about the infrastructure quickly, as I said, there is the environment as this production draft and public environment split. Then there is VRED Server. And it's also API applications. I will not talk about the API applications today, because all what we are going to do is in the back end or in this interface. A API application would be a additional front end, like a configurator data front end or something like that.

Then the repositories, we've discussed that already. If I click on them, you can see that we have the repository for CPU-- that was also something we tried-- a GPU, and we got a local repository. So this is really just to register the notes that we have. We got a image repository and the scene repository. As I said, I won't go into detail too much on those, because it's not too interesting for us as users of collaboration sessions.

And there are some monitoring features. OK that's just that part but also have some error logs that are updating in real time. Auditing section and a media part. And this is where you could access your media. So for example, I've uploaded a document, which is VRED Server documentation. So people could download it and use it.

Then there is the image repository content. So let me just do that 48 so we see a bit more. So you will see in a second that these are all the previews we have for the interface. If we are generating images on demand, for example, then those are stored there. So all the images you have are stored there.

So maybe we are lucky, and there should be also the profile. We've generated too many images. Because there are also the profile images of our developers. But I don't know where they are yet.

And then there a place to store scripts. You are able to apply a post- or pre-Python script to file when it's loaded onto a note. And you could store a set Python scripts. For example, this one, we could take a look at that. It's quite easy. It's just giving some instructions on the render settings. We also have access for the render settings in the tool, but just to give you an example of what you could do that.

You can also provide, for example, an automatic preparation script, which applies shaders or something like that. So you could be creative for that. It's working exactly the same how you know it from the command line way in VRED Professional.

Then we got Settings. And you will see that there are a lot of Settings. So we've structured them in different ways. We've got Collaboration Settings. We've got Mail settings. So for the Render Scheduler, for example, if your job is done, you will get a notification where you could find your images. But you need to register that, so yeah.

And network settings, render job settings, specific scene settings, and user settings. Not too interesting for us right now. Just to give you an overview of what's possible from, let's say, an admin's standpoint. Also talking about administration. And I think that's something interesting to this use case. This is the users.

So as you can see, there are some users pre-assigned already. Most of them are our developers-- so Mario, Bogomir, as well as my account. And you can see there are different roles assigned to it. So as I said, I'm an administrator. So my interface looks very complicated at the moment. I will show you how it looks for a user in a second. And you will see that it's way easier for those people to use it.

Yeah, so the structure is, you would have a user, you would assign a role, and you could define a role by assembling different grants, which are basically access rights to certain functions of the back end. So if we take a look from the bottom to the top, a grant could be create animation groups, for example.

So I could assemble those 2, as I said, to collaborate, for example. I'll produce that a lot. So they are just able to join a collaboration session. That's it. But you could do whatever you like and give more access to people you think able to cope with it or able to handle it or to just different types and personas. Yeah, so let's take a look at the roles. We'll take a look at the collaborator.

You can see that the ground groups below here are completely unchecked, besides the user. User settings-- so this user would be able to take a look at his own user profile, while in the administrator role, there's like, a lot more. So you can really easily assemble those for different people in your departments.

And the last point on the interface, and then we're going to talk about how to create a collaboration session, is my account. So my profile with password and details like email and stuff. And then the profile, which we already saw when I clicked on that one before. And the last point in the menu-- an info page, where we are listing very interesting stuff about third party stuff that's used in the tool. And some other disclaimers and legal stuff.

OK, so that was it for the overview. Now we're going to set up a collaboration session. And as you can see, one is already scheduled. But I will set up a new one. So I will go to collaboration sessions, and go to create collaboration session. Before that, we need to double check whether the node groups are configured correctly. So I will do that. I've set them up so it's-- I'm quite sure that it is running-- but to show you how it works and what it is referring to.

So I will just give it a name. And it needs to have a role, and I think that's a very important point for the node groups. So you need to pre-assign assign a role-- what the nodes are allowed to do, what they should be picking up if there is a request coming in. So you could define them as a collaboration session node. So whenever I'm starting a collaboration session, this node group will be used. You could assign random jobs to it, so it would process the random scheduler queue.

And you could define preview, which I'm going to show you in a second. And scene is something related to configurations. So this is on-demand rendering, which I don't want to talk about today. So we will leave that to collaboration session. I will not assign the scene, because the scene will be coming with the collaboration session. So that's flexible.

Here you can see where to upload your scenes. So either you can drag and drop just-- not scene-- your Python files. So either you can just drag and drop a Python file for pre and post Python onto that image, or you could load it from your repository. As you see, this is the rendering script that we just saw a bit earlier, but we don't need one for that session. So I'll leave that open. I'll enable it.

And something that's interesting-- I will have just assigned one node to it. But just to show you how it would work-- so you have all your repositories listed on that side. It's not too much for me for my demo. But that would be all your nodes listed. And you could easily assign nodes to a node group just by drag and dropping it over. In this case, it would create two different streams. I could also cluster that, and assign like, I don't know, how many nodes you need that goes to it. That needs to be in local. That's the wrong one. I think I'm just messing it up.

Let's see. The good thing is, it's always jumping back to the right repositories. So I've just flipped the IPs. But if I remove them from the node group, they will fly back to the repository where they belong to. So yeah, that's it for the node groups. As you can see, I've also set up something completely equally besides the role-- it's preview. Yeah, so that's the second node group just consisting out of one node-- one GPU node. And its grid nodes, so it's not a very happy computation unit on the AWS.

Users-- I think we took a look at it already, but let's go to administration. And we could create a new user, but I don't want to at the moment, because I've already done that. Let me extend that to 20. And you can see those are the 10 test users that we've created, as I said. At the very end of the presentation, we could try if you like to join a review session with your cell phone, for example.

Let's just take a look at how a user looks. Yeah, so the difference to my profile is just the role in this case. So we already took a look at that. There is a preferred language, in this case, English. Yeah, and it's quite easy to set up. So registering those users at the system is a very easy thing to do.

As soon as you have assembled your rolls, you can assign it obviously to any user you're creating. So talking about previews and scenes all the time-- let's dive into the scene section. Again, I will not create a new-- let's try that. I will create a new scene just to show you how it works. So I'll have to choose my repository first, if I'm creating a new scene. I could either upload a file to this repository. So I need to choose it first to create something elsewhere to put the file to.

And in this case, I don't want to upload it, because I already said that the internet is not very fast, so putting up a 2 gig file would take too long. So I can just browse from our repository. And let me check-- I can use that one, which is basically the genesis, but just to show you. Test. Let's-- here we go. Genesis AU tests. And we could also add a description. I'm not going to do that now. And click on Create Scene. And you can see that this scene was successfully created. That's what the system gives back to me.

And you can see some animation groups, varying groups. This is what we know from VRED. This is the content from the varying said editor, for example, module, for example. There are some sequences, there is settings, and there is the Preview button I'm talking about since quite a while.

So let's just open animation groups. For example, you can see there's nothing in there, right? So also for variance, there's nothing in there. This is because VRED is not automatically loading that from a file. We need to tell him. And actually, VRED servers, not even loading and opening the file, loading the scene configuration. But it doesn't even need to open the file, because we have attached the same configuration as an extension to the DPP file. So we can just click on Load configuration from file.

And I think this is going to take, I don't know, 10, 15 seconds-- so it's some time. What happens is it's accessing the extension. It's reading out stuff like, the previews for your variance. So those are generated automatically in VRED in the variance set system, in the variance module. And it would just read the whole scene configuration and applies it to the file. So we don't need to set it up by hand.

You could also do that if you're referring to the right variance set in VRED, this would work, but I think it's a very annoying thing to do. So I would try to automate it as much as possible. Here we go. So it was a bit more than 15 seconds, but you can see now, the preview of the scene is there. If I unfold variant groups, you can see there is a lot of content now. Let's take a look at car paints, for example. This is a variance set in VRED, so you can see, its variants. And if we click on the variance, you can also see the previews we've already created in VRED. So those are also automatically created and read from the file. And I think what's important here is, you can really alter what's later on visible in the collaboration scene.

So if I go to the upper right and click on that button, I would be able to restructure or move the position. And what's important here is, the first one in every category will be the one that's executed on start. So if we're starting the collaboration session, in a second, the color should be black. It should be in an environment in the-- outdoor stage I think is the name of that environment.

If I don't want to, like, in this case, I really prefer doing the studio as the start. So I would just drag and drop that up there. And you can see that studio is on top. And then, if I hit on Start Calibration Session, it should also use the studio. Here we go. Those are very specific settings to the scene. So there are general settings as we saw them already. There are also scene specific settings, which will override the general settings. So for example, if you set up general render settings, and you set up different render settings in your scene related settings, it will override them. So the scene always overrides the general one.

And you can see my personal friend here, which is the bit rate of the streaming codec. So I've reduced that a lot to get them running. And now we're going to-- so if I'm fine with all my setup of the scene, I can click on Preview. And now what VRED does-- and I can show you that in the Dashboard-- is, here we go. This is the preview node group. It's loading the data set. You can see it's just firing up the random node. It gives us the feedback on the status.

And as soon as it is ready, it will provide a stream here. And what you can see on the right is already the interface of the final product. So the collaboration session interface will look quite the same as this one.

So just quickly, let's just wait. Should get that running. Here we go. And it's executing an animation. So it's starting with the configuration I've setup. Yeah, because I used that file. OK. You can see below that you got all the configurations with previews. So I could go to car paint and switch car paint, for example, as you know it from VRED. There's different environments and all that. But we'll take a look at that in a second.

So this is really just to check whether your configuration you've just set up is correct, whether it's fine to, let's say, use that scene for a collaboration session. And if we decided to do that-- let me just check. So this should stop in a second, because I've stopped the preview. Let's just go into a collaboration session that was done already. So I would just start it up. And again, this time, it's using the other node group, or it should use the other node group. Yeah, here we go.

So it has removed the preview group again. Now it's loading the file onto the collaboration group, which I've defined for my collaboration session. I just closed it. Collaboration. He we go. And it's starting. And I can just click on Join. And again, VRED is loading the file onto the render node. This will take a bit of time, but we can go to the interface already. So I will reduce that one.

As you can see, this is like the preview. You have your render window with your VRED content. You have different variance, different groups, variant groups, like you set it up in VRED with interior environments, car paints, license plates. Yeah, so you could imagine what you could put up there. You have viewpoints, so different viewpoints. And by the way, the whole thing is compatible with touch, obviously, if it's running on an iPad.

And you got bookmarks, which are basically situations in a review that can be like snap shot-- can be stored. So if you, for example, if you find a configuration that you really like that you didn't think about yet, like a car or something, you could just hit at a bookmark, and then bookmark is edit. And you could get back to that anytime. So this is like a variant you create in VRED Server Manager. So if I click on that, I'm getting back to the bookmark that I'm coming from. However, that will started previously.

Yeah obviously, as you can see, this is a dynamic stream. So I'm able to either like that, or also with touch gestures, I'm able to use that. And you can see some compression artifacts because of the network. But I think it's better than when I started the session, so that's quite good. Completing the bookmark piece-- if you decided to save a certain point of a discussion, for example, you can also annotate to that or put notes to it. And we did that already. So you can see, this is the bookmark with a, let's say, we choose that one and take a look. And there is a comment that obviously I wrote.

So this is a bit like egocentric demo, because nobody else is collaborating with me when I prepared that. So I can leave some messages for people that are joining the session later, or we could just write down our findings to have them for the later process. And on the right side, you can see, there is the bookmark-- node section already. Then there is a list of participants. And now I'm trying to join with the iPad. So you can see that somebody else is joining there. So the collaboration session.

And this is what I meant. Unfortunately, it's quite small, but you can see on the right side that the menu is way smaller, right? So this is a collaborative view. I can show you that on my screen as well. What I'm doing here is, I will just click to collaboration sessions. Then I get completed collaboration sessions upcoming and running. I will choose the running one and just join it. Here we go.

So the scene is already running. And obviously, this is design collaboration, which means, we are discussing the same thing, just one stream. Not everybody is doing what he likes, because it wouldn't make sense, right? So if I'm using-- I'm at the moment, the presenter. And if I'm moving it, it's also moving on the iPad. So people will follow me, but we can hand around the presenter functionality, so if somebody else wants to take over, I can just click here, give them the presenter functionality, and then I could use the iPhone or the iPad and I'll leave it here, or, I don't know, change variance, whatever you like.

So getting back to me, because I like being the presenter. The second thing is annotation-- scene annotation. So something that's just relating to scene, not to bookmark, not to a certain state. So for example, if I have some information for an operator, for example, please change the number plate, or I don't know, something that I don't like for the scene, or for the next review that they should change, I can just add an annotation. That was the wrong thing. I can just add a note. And it will be just added to the scene.

I will show you how you can access that without accessing the scene in the second. And the last thing that I want to show in that collaboration or preview is the 3D annotations. So we have worked on that. Those are computer compatible with the VRED annotations, so you could read it out of the VRED file. But they are more web based, right? So we have a bit more flexibility on creating them. So you could also market in 3D quite easy.

So if I'm creating an annotation just to show you, and I don't want to do anything too meaningful, I can pick a position. So for example, the logo of the car. And you can see, there is the new annotation. I'm able to choose whether it should be near or very far from the point, so it's not occluding what we're discussing. And then I can just save it. And you can see, it's following the car.

If it's occluded, it's going to be like, the light grayish, and finally disappear. It's behind the car. OK, so that's for the session. So maybe to say something general-- the session could be done simultaneously. So I could invite 10 people to join me with the session, but I could also just open it without the timing and just leave it open. That's what I just did. So we just joined a session that was officially already done. And people could just drop in, leave their notes, drop out. Somebody else gets in, like leaves a note. And then, finally, this is what I meant with the integration, or let's say the conclusion of the whole thing.

If I'm going to that running session, you can see a small icon here. If I click on that, I can get a summary of everything that happened in that session. I could also download the SPDF, get it to an operator. He's changing the file, uploading it again. So you can see, I've joined that session quite often while preparing this presentation. The bookmarks are stored there. So you could get them as a PDF with the annotation, with the bookmark annotation. And yeah, some details to the session on the right side, all the stuff that was marked or annotated on the left side. And as I said, it's also possible to export that.

It's also possible to export the 3D annotations to VRED by XML. I'm not going to show that, because we're running a bit out of time. But you could just-- so I've just created that 3D annotation. I could just export it via XML, attach it to the PDF, put it into an email, send it to the operator, he will read that into VRED, alter the scene, sends it back, and we can do the next review based on the altered scene.

And I guess that was basically it. So for those who would like to, and I guess, as I said, the internet connection might be limited to you. If you'd like to join, there is the IP. It's upgraded users from 1 to 10. And the password is user1234. I would just have that running, because, as I said, we are reaching the end. So I'm happy to answer any questions that you have. I will hand over presentation mode if somebody of you is logging into the session so you can play around with it. And yeah, with that being said, I hope that was an interesting presentation. Thank you for your attention. And as I said, if you have any questions, feel free to approach me. Thank you.