DigitalituM Podcast Episode 3 - Marek with Aidar Solutions- Technical Training in VR

DigitalituM Podcast Episode 3 - Marek with Aidar Solutions- Technical Training in VR

Summary

 Marek Carbon, co-founder of AIDAR Solution, discusses their VR and AR training solutions for manufacturing. IDAR Skills is a no-code editor for VR training, allowing customers to create and customize their own VR training programs. IDAR Service is an augmented reality application that provides step-by-step machine maintenance and repair instructions. The solutions are hardware-agnostic and compatible with various VR and AR devices. They also offer remote support and IoT integration. The goal is to bridge the skills gap in the industry and provide efficient and effective training solutions. The biggest challenges during the implementation phase of the IDAR and IDA solutions include teaching people how to use augmented reality, the lack of standards in AR/VR, and the difficulty of creating interactive 3D models. Implementing the technology's cost and scalability concerns can be addressed by using the IDA service for step-by-step tutorials and starting with more straightforward AR solutions before moving to more complex VR training. The future of immersive technology lies in integrating AI, such as using AI to generate 3D models and analyze user movements for ergonomic analytics. The lack of Wi-Fi connectivity hinders the potential of IoT in the field, but advancements in satellite internet and improved security measures can overcome this challenge. Some use cases of the IDAR solution include training employees at Daimler's new plant, remote support during the COVID-19 pandemic, and training dentists on 3D scanners. Implementing immersive technology is recommended now, as the hardware and software systems are mature and capable.

Sound Bites

 80% of the people who got the VR training were able to execute the procedures directly on the patient with very little assistance from experts, and no one was able to execute the procedure from the PDF, just from reading."

"You have an integrated video inside of the tutorial. And this is, you know like the cheapest or more with, you know, like every phone, not all of them, two or three years can support this kind of stuff."

"You're also independent, as you mentioned, from a timeline, you can simulate faults. And you do not have to wait until that specific situation appears."

"What are the biggest challenges you face during the implementation phase of either solution?"

"It is a challenge to teach people how to interact with this kind of device."

"The ROI is not there. How do you challenge your potential clients that this is a good solution?"

SHOW NOTES and LINKS:

Digitalitum Website: https://digitalitum.com/

DigitalituM LinkedIn page: https://www.linkedin.com/company/digitalitum/

Markus's LinkedIn page: https://www.linkedin.com/in/markusrimmele/

Marek Carbon LinkedIn: https://www.linkedin.com/in/marekczarzbon/
Aidar Solutions website: https://aidarsolutions.com/

Stay tuned for more inspiring conversations about manufacturing and digital transformation. Also, remember to follow and subscribe to the DigitalituM Podcast for exclusive insights from industry leaders and innovators.

We appreciate your likes and comments. If you feel you can add value to this podcast series and want to be our guest, send an email to Sales@DigitalituM.com

Transcript

0:03

Welcome to the Digitalitum Podcast where we delve into the intersection of manufacturing and digital transformation. My today's guest is Marek Carbon, co -founder of AiDAR Solutions solution from Poland. Welcome Marek. Welcome Markus nice to have the speech with you. You are welcome. So let's jump right into it and tell us a little bit about your background and how did you become co -founder of AIDAR AIDAR Solutions What's the story behind that? My background is like I studied computer science in Germany. I studied at the same time also graphic design. So the combination of visual effects, of visual aspects and computer is something which is in my DNA. And this is like, I don't know now, like eight years ago, I was running a company which was... dealing with SharePoint as a software platform. So I was Microsoft partner and then they come to conference from Microsoft telling about the HoloLens and the possibility of the HoloLens. And it looks very spooky, like what is that? But I decided, okay, I will show it out. I just register for the beta. and we become one of the first European partners of Microsoft with this HoloLens. And after trying it out, it was like, you know, a little bit strange, but still good enough to understand what is the idea. And for me, it was clear this is the future of, you know, human -computer interaction, to have the whole room, to have the understanding of the room. And at the beginning we started with, you know, like small vision application for the health and years for the systems, this kind of stuff. But then it started that we realized that we have like repeatedly the same stuff which we are doing and that it makes sense to make a product. A product which will allow our customers to adjust, change the training which we are doing, the reality by themselves. So this is how AIDAR Skills, which is like no code editor for VR trainings came out. And we started with for DAIMLER which was building the first company or first plant in Poland, in Jawol. And they have the goal to train the 1 ,500 people without having the actual manufacturing plant there. And this is how we decided to make a digital twin of that. It was like 80 stations, every station 70 seconds, and we trained the people inside of that reality. So this is the story how it started with AIDAR And we decided to find an investor so we can readily develop the product because at the beginning you need to you know, invest a lot of money to have a stable, nice product. And now we are already in the second round, you know, gaining more customers and very interesting product, project. Very good, thanks for putting that all together. Yeah, let's dive a little bit deeper in what is actually AiDAR Solutions solution? How many people do you have? What are your core focus areas? What general problems do you solve? Maybe you wanna show a quick, your company presentation. Sure, would be nice. Give me a second. I just share my whole screen. Cause I have a combination of some screens and some videos. Okay, let's go to the full mode. So maybe first like why either? And it was on the beginning already the combination of artificial intelligence to support the... vision, object vision inside of augmented reality of the HoloLens and using of augmented reality are and also the English word aid to aid somebody, to help somebody. So the AIDAR is the helper, uses the technology to build something like a knowledge base. So we want to build a knowledge base of instruction and especially now it's very important. You see like we have in Germany, in Poland, I know, I guess in US it's the same. The problem was how to transfer the knowledge that the young people are not really coming to this kind of jobs. So you have to make it more attractive, more digital. And this is where we are stepping in and making this kind of solution. And as I mentioned, we worked for Daimler, for ABB, for Canpack. Kallpark is, for example, a company which is producing aluminium cans for Coca -Cola, for beer, and they're producing 5 ,000 on one machine per minute. So it is, you know, if you stop the machine, you have to handle it very quickly to repair it. And what we do in virtual reality, we have like a simulation of a failure and the service people are trained how to manage it. So in case it happened, they can react very fast. And, you know, like an anecdote is like before they have to send the people for six months to another place. Why for six months? Because the probabilities that this kind of failure will appear was low. So they have to stay there for six months just to have the probabilities that this every case of failure will appear on the production line. This is something we are running on OpenXR so we can run standalone VR glasses. We can... also to like Unreal or HoloLens or Magic Leap. So you will see in a second we have that boss solution for that. And what is our typical? Yes, yes, sure, sure, sure. We try. whole immersive technology space from AR to VR to XR. exactly. The whole scale. Okay. And you know, like we have AIDAR skills and AIDAR skills is the snow code editor for VR trainings. And for us, the VR trainings are important if you have something like a high risk, the machine on which you are training people are very expensive, is very expensive to stop the whole process. So in this case, it makes sense to invest the money and, you know, have the 3D models of the machine and create the trainings. And it was the no code editor. Usually we help the customer at the first phase, but then we handle over to the customer, we train them to do it. And if they have like a changes because, you know, a lot of the company manufacturing are doing the six sigma and lean management. So they're making like a small changes, but very quickly, like every two months. So it is very important that they can change the training that is up to date with the actual process which they have. the owner of the training solution can make the changes themselves without reaching back to AiDAR Solutions to let them make the changes. And going really quick back to the basics of training. What I see as a problem out in the industry, specifically in the area of machine building and maintenance is that machine. Machines and systems get, get more and more complex. That's the one end and the skill level out on the, on the shop floor kind of stays stagnant. and that creates this gap and the traditional training methods in I'll, I'll take a junior with me and, and show him how to do it and kind of shadowing that process, which then takes years until this person gets all this knowledge. that just takes too long. And then specifically in the U S we have the problem that people come and go to the workplace that they don't stay like 10 years in the same company. And I think in this particular problem, that's where the training solution comes in. And the virtual reality has the main advantage that you take a cat model. So you're independent from a location. You're also independent, as you mentioned, from a timeline, you can simulate faults. And you do not have to wait until that specific situation appears. And so all this training can be done already before and maybe on a basic level. And then only the final touch you do really on the, on the real machine. Do you see it the same way? Exactly. For sure. And it is also like, because I question it like, you know, you do this virtual reality training, but you don't have the haptics, you are not ready, able to feel, you know, the force back or something like that. But our brain is constructed this way that we are memorizing the information for a long time in another way. We need like a picture. This is why it is important if you learn even from the book. It is a study which was run by one of the big consultee companies and they were run with doctors. They had two groups. One of the groups get the VR training and the other just got the book with description of the procedure. Mm -hmm.% of the people who got the VR training, they were able to execute directly on the patient the procedures with very small assistance from experts and no one, no one was able to execute from the PDF, just from reading. And this is how our brain is working. We need this picture from different perspective. The brain is making like a lot of pictures and this every picture is like connected to information and the land like if you learn something you should for example read it first time you know in the kitchen and the second time in the main room because you need like another surrounding so the brain is challenged to save the information in another with another picture with another information and in virtual reality we can do in the perfect way because we can change the lighting which can change the surrounding, we can change that you memorize easier. And we have the one -to -one scale, so it is like the muscle memory because you do exactly the same movement. And for the brain, it's not important if you do this movement three times or only one time, you just recognize or remember that you have to rotate. You have to move that from A to B. And this is enough just to have the indication. And as you mentioned, Markus, before, then if you're on the real machine, yes, you are still have some gap. So you have to learn, but it is like you are on another level. You learn the difference very quickly then. Yeah, you have done it before and you already kind of understand the environment. And that's where I see the virtual reality training or the immersive space training step up in, let's say a step down is watching a video in how to do it. I guess that's the step between reading the books and the PDFs. That's the lowest level. Then the medium level is you watch a video and then they have currently the highest level is you do it actually in immersive space. As there you have the combination of what you see and as you use your hands to do things, you get this whole muscle memory into the play. Correct. Okay, but. Yeah, just continue. and AIDAR service are saving all the data in the same database in the same structure, in the same format. So we can change between one and another one. So we can start in augmented reality, go to virtual reality and back. But here. we use only one function of, you know, augmented reality and this is the special understanding. So if I have here a glass and I put my finger here that I know this is on the glass on the top. because of that, but I don't need the glass because of the real glass. We just need the calibration of the glass so we can do it. So here the application, we are doing the step -by -step workflows, processes. For every step you can add like a video, you can add picture of pictures. You can record your instruction because to write something on the HoloLens without this force pack is not very convenient. And then we convert the spoken word to text and you can correct it. We can also connect it to IoT signals because like if you have a machine with IoT and instead of going to the display and going deep in the settings to find one of the settings, if you have an IoT connected machine, you can expose some of the settings to IoT and then we can include this in the instruction. So the data is showing in front of you, showing that you show an actual value, a temperature, an amperage or something at a specific location. As we see here, a video was a person wearing an Yeah, this is the HoloLens. Yeah, and this is in front of CNC machine. And in the second, you will see, and for people who don't see the video, just I try to explain it. Here we see the interaction with our application. So we have in the front of us like a step list, and then we have something like a... bubble with information about the current step, but the current step is shown in the screen. And then we know where to position something. And in the second, you will see also the IoT signals. Okay, I just, okay. So I just stop here. Oops. I might just run back in the video. Okay, so in the screen we see like two signals and the signals are coming from IoT. In here is the angle of the axis X and then we have the number of pieces produced and this is something which will appear in front of you if you're in the one of the step eight which you see in front of you and then you can follow. And this is the combination, you know, like you get the most... important information in front of you. You don't have to look at your phone. You have free hands and then you execute the steps. And after you have done it, I just show you now live. Okay, give me a second. Just go to the server. And now I'm logging to our web platform. We can run multiple tenants in the same platform. We are delivering as Docker, so you can install it on premises as well. But let's go to this training which we have seen before. And here you see the Hamla CNC. So I go and click in the training. I see the training in the web application and one of the lesson. And now I am showing the lesson. So you see here. It is flat, it is on the screen, it is responsive, so I can run it on the smartphone, on the tablet. And the idea is like you click here with your thumbprint and then you go from step to step to follow if you don't have the augmented reality glasses. Mm -hmm. But still, if you're on the flat screen, we have the special understanding. So we show you where to do something on the very simple hologram of this process. And for every step, you can decide what is the most important. This is a video, so this would be the first one. Here, for example, it's a video. Or the position, so we still have the hologram. And then it is easy to change the text. The text will automatically read aloud. Mm -hmm. And even if I change the text, it will automatically, you know, with artificial intelligence translate with neural language, neural voice spoken so you can follow the steps. And just for quality control or something like that, we can export all the manual to the more traditional, you know, like step -by -step PDF document where you follow the steps. Also for audit quality assurance. Or we can go in and you will see in a second and see it in augmented reality. So, if I go back. and go for example how it would look like on the smartphone here. So here is a screen on the smartphone, but in augmented reality mode. So you just keep the camera in front of you, the phone in front of you and you calibrate it. I will just make it a little bit faster. So you see in front of us an engine and the engine, you have like a calibration picture. So you'll need like two characteristics points of the engine. And then we start the lesson. I will do it in a second. Start the lesson. And now you can follow and you see even if you move the phone, where is the point you have to do it. You have like integrated video inside of the tutorial. And this is, you know, like the cheapest or more with, you know, like every phone, not all of them, two or three years can support this kind of stuff. Yeah. And sure. the last one you showed, how long does it take to set this up? Okay, this is, I will say something like, if you do it like the job normal, for example, for two minutes, then you need like, like factor three times longer to do as with as instruction. Because you just, you know, you have three hands, you do it, you just add some pictures, you record the voice. So for example, I would say something like between 10 and 20 minutes for instruction, which have something like 15 steps, something like that. So very quick and you don't need the hologram. very good ratio because the challenge on a lot of this technology solutions is that the time it takes to create the content is way, way, way longer than the actual use case. And that hinders people to create the content. In addition, when it's complex to create the content, for example, when you need a full scale CAD model with all the details, which often is not available, then you cannot go that route. But in this case, what you just showed, you take your phone, which nowadays 99 .9 % of the people have a smartphone in their pocket. And by connecting to the platform, you are able to create content. Exactly. That is just phenomenal in my point of view. Yeah. It is exactly what we get from the customers because we would try to convince them to go for the VR training, but we realize that sometimes it takes too much time. And if, for example, you have somebody who is delivering machines and they have like only one machine for the customer. So the effort to make a virtual reality training for the one machine is too high. Yeah. But if you can do it in front of the real machine before you deliver to the customer and you spend maybe one day. for 20 instruction. And then you send it and then the customer can use it. And we are, which is, which I didn't mention before, but we have also in our tool included remote support. So in case you are not able to find a solution, we can still have connection in the same glasses on the phone with the expert with some annotations 3D, which you can do. So exactly, exactly. So it's actually the combination of having the instructions in augmented reality on the device you have already in your pocket to self -service. And if you're not able to do self -service, you're still able to connect to an expert and wherever that expert is from a geographic location. Exactly, correct. That's pretty cool. Yeah, and what also we did is something like, for example, we have customers in Japan and some in Europe, so it's like big time difference, and also language difference. And you can send a video with the problem, and then somebody in Japan can record very easy step -by -step instruction how to solve the problem. why Japanese, then we translate automatically to English or to German or to Polish and then somebody on the next day can follow the steps on the exactly the same machine and solve the problem. So which makes like 60 % of the cases easier. Very cool. You're sharing right now the slide about the device agnostic. So more or less, Ida solution has this platform and which whatever hardware you have, augmented reality, mixed reality, headsets, all the different virtual reality headsets from the Quest to even I see the Apple vision pro on there. You guys cover it all. Exactly. Yeah, Vision Pro we ran to Munich, you know, before it was already, you know, on the market and tested it and after eight hours of, you know, like adjusting it was already running. So which is very nice that we know that, you know, we can adjust to every device supporting OpenXR and Unity because behind we are using Unity, but our application are standalone applications, not a plugin for Unity. and then on top of that is the standalone AiDAR Solutions solution. And from there you can go to whatever hardware you want to go. Yes, usually we need some adjustments, but it is not a big problem. So we test a new hardware, new versions. And so even as hardware is developing and new hardware is coming out, you're kind of always ahead of the game and have everything prepared. Yes, and this is part of the service of the product license as well. Mm -hmm. Very good. What else have to have you to show us? This is like the whole picture of what we have. So what I mentioned before is AIDAR service, this is easy step -by -step instruction. Then we have this AIDAR editor, which is doing the, you know, we call it full -blown XR training. So you don't need the real machines. In both, we can do metaverse or in either skills, if you create a training or you just have a machine, we can do up to 10 people together in one room and you have your avatar. So it is for the trainer that the trainer can explain the machine, the parts. We can also run the instruction in Metaverse. So somebody is the presenter and showing like how to do it and the trainer can check it and see it. As mentioned, we have the remote support for that. And then we are starting with IoT connectors and device template. Give me a second. I just, sorry. I just run here in the device templates. Okay. So the idea here is have, for example, production place site, and then we split the whole plant in sectors. And in sectors, we can have different machines. So we have device templates, which is like a class. And then we have the devices, which is like instances of the devices. So you can say in which part of your sector is which machine. And if you're front of the machine, we are showing your instruction only connected to this device template. We can also connect with IoT, so we can create like... IoT signals and mentioned before and if you're front of the machine you can you can see it but they are we also use for predictive maintenance and what I mean like we don't have like predictive maintenance included in our software but if you have predictive maintenance and predictive maintenance is saying you have on this device to change something, we can say, okay, if this warning is coming, here's a tutorial which is connected to this warning. And then we have the step -by -step instruction how to do it. And this is addressing the problem like you don't have the experts or you are executing this kind of procedure not every day, but every two months. So you forget it, it's natural that you forget it. You don't have to recall it from the PDF and figure out, this is from behind or in front. We just, one time you have to set up, but then you just follow the instructions with you, which is much, much easier than to follow it from PDF to comment. That is pretty unique. It is. Okay. Good. All right, kids issues are solved. Okay. This is so to sure. All right, very good. Thanks for giving us that overview in the different products and solutions you guys have from AiDAR Solutions Service and AiDAR Skill. Just as maybe some of our listeners got a little bit confused, what is AiDAR Skills and what is AiDAR Solutions? Could you break that down again and kind of explain it? sure. on the one and what is to focus on the other. Sure, I guess you mentioned AIDAR, you wanted to say AIDAR service because AIDAR solutions is the company name. Okay, and AIDAR skills is started with virtual reality. So with the full blown... full holograms, interaction, all the stuff simulated inside of our application and either serve as something you are in front of the real machine and we just know where you are, you calibrate on the machine and then we show you like a checklist step by step what to do. we don't need the whole holograms interaction animation because you are just on the real machine. And there's integrated this remote assist. So you can call an expert and have like a 3D annotation. or you can be in front of the same machine, but a different location, and you see like an avatar of another expert. So the conversation showing something is much easier to support you. I hope this was clear enough. summing it up again. So the one is the virtual reality application and the other one is the augmented mixed reality application. And both technologies have their pros and cons and their different use cases as with virtual reality, you're... enclosed, you are in a virtual reality, you're independent from the location and you can, for example, train on an equipment which is currently running or at a geographic different position. While with augmented reality, you get the information you need at the location you are in order to do your job right. That's... future looks more like it would be like the mixed reality. So you see on the Vision Pro, you have like something you can just turn and say virtual or augmented reality. And actually for us, it's the difference you are in front of the real machine or we simulate the machine. And this is the difference between either skills and either service. As you have the solution out there and implemented it already at a lot of different customers, what are the biggest challenges you face during the implementation phase of either both solutions? Okay, the AIDAR service, because it's easy, you can start actually the next day. So the first you have to teach the people sometimes how to use argument reality. Because we have Windows or Macs for years and we know how to use the mouse, how is the window, how I close it, how to make it smaller. the standards in augmented reality, virtual reality, they're not standards at the moment. Yeah. So you have different kind of. how you call the menu, so something may be on your hand, sometimes it's in the room, but if you move, what about the menu, is it coming with you or not? So there's a lot of challenges, and it's more like you can compare it more with the real life. So if you go to your TV, in front of your TV would be some buttons and then you know, okay, I have to click it, but the buttons don't follow you. Mm -hmm. but in the digital world, we can do this kind of stuff. So what I wanted to say is like, it is a challenge to teach the people how to interact with this kind of devices. You know, if you have a smartphone, exactly. Yeah, exactly. And. Okay. And everybody has at the moment smartphone, so you can accept a high level of already induction. And so you see kids and they see like advertisement on the streets, they want to touch it and whatever. So it is assuming more that everything is digital. But if you show somebody HoloLens, there's not very common that people have this kind of device at home. So for them it's like very new and... They are more like I have the challenge like how to interact with this kind of device and then they come into your application and they try to understanding what is your application about. So this is a challenge if you show it to new customers or to somebody how to do it because you have to overcome a lot of obstacles in the learning process. Mm -hmm. And this is difficult sometimes to explain. Okay, but this is like interaction with the hardware. And then the process, how you create a virtual reality training. And for example, if we say, I don't know, you want to create a new website, so you need that CMS program, CMS. So this is clear for everybody. And usually the company has something like a marketing department and the marketing department is... responsible for this kind of stuff. But what about virtual reality? What about, you know, special understanding, whatever? Nobody there. So you have to find actually a person who has time for this kind of... Exactly. Yeah. out and work on it and put their emphasis in to make it a success. Yes. So if they say we want to do it, they have like a big exposure because nobody's expecting that this will happen or you use this kind of tool because this is like still like an innovation. So you need the acceptance of your, you know, your managed boards that they're accepting it, that they come in with this kind of stuff. And then... we'll start the project and the next part of how I get the 3D data. Ideally, it would be some, you have the CAD data, CAD data and you can start with that. But it is still not common that the producer of the machine are giving like a reduced form of the machine as a 3D model, you know, because I understand is that they are not wanting to, yeah. Yes. Yeah. data gets stolen and copied and all of that. but they should, you know, they can export something like a reduced form of 3D data of the machine, which is fine enough. The glasses are not able to, you know, render all this data. It doesn't matter. Yeah. And it should be like a more like a political movement or something like that, which will, you know, force the people or at least like give you like a acceleration for this kind of the producer will give 3D models of the machine just to make it. Possible, yeah. But a lot of cases not. So then we have to, you know, like make pictures. So we going for the digital world back to very analog one and we make a lot of pictures. Then we hire 3D graphic, which we're doing this 3D stuff. And then we have model with the similar. So this is something which is difficulty. And then. even your Flex3D models, you have to say, you have to make it interactive. So it's like a, where you rotate something, is this part a separate part in the 3D model? Because it's not separated, it's not so easy. So it is a lot of, you know, puzzles which you have to have. or to solve to make a VR training. So the easy one is like we are cooperating with the producer of the machine. Then it's still not, the lower it was very easy because we did, we get this card data. And the only challenge in the card data is the size. Is there a lot of polygons inside? And if you have, for example, a skew, every, what is the English word on skew? You have this, I don't know what is it, gewindel? threat. Yeah. this, like, everyone is like a polygon, yeah? So it's just like, you get already millions for very small part. But now we have like a lot of Python scripts, which we run in Blender, which automatically reducing the number of polygons. So for example, last week we got a card from the customer. It was like 5 million polygons. And after... half hour of some standard operation, we got it down to 900, yeah. This one million polygons are already running, just to finish the sentence, on the Quest 3, it's standalone. And this is nice. This is something you can do some. Yeah, and just to sum it up again, specifically about the polygons, the issue is if your model is too complex and too detailed and has too many polygons, then it takes a lot of hardware resources on the virtual reality side to actually visualize that. And if you turn around with a large model, things are flickering and do not look. good anymore. And that's the reason why the data needs to be compressed and all the unnecessary details need to take out. And that's the process you explained, you go through. And also to sum up what you said previously about the challenges. So it actually starts on the people side of things. People need to be ready. It's a new thing. It's a new task. So people need to be willing to do that. There's also the leadership challenge that leadership needs to support that. Then of course, always the issue with CAD data. You have to have the CAD data to do something. If you take back pictures or maybe even a 3D scan is a possible way as well, but then you get this point cloud and still. That's one piece where you have to piece mill things out, which is again a time intensive process. Did I sum that up correctly? Perfect. Very good. And then from there, we kind of get right into this how to address concerns about cost and scalability of this new technology. Because when you just start on this, this sounds all very time intensive, very complex, and that translates normally directly into very high cost. So therefore, It doesn't make sense to do it because the ROI is not there. How do you see that and how do you challenge your potential clients that this is actually a good solution? exactly. I mean, this is what I mentioned in the beginning. This is why we have AYA service where you can create step -by -step tutorials. And they are very cheap to create. Mm -hmm. And also to adjust, to change, even if you change part of the process, you just one step you are moving to another part and this is ready. And this is what I guess it's good enough for 80 % of the cases. You don't need more complex stuff. And we can, you know, to every step you can add a picture and video. And why? You know, you can do what will be the alternative. You can do like a PDF or you can do a live video. But the problem of the video is like one long video. If you change your process, you have to find this in the video, change it, and then do it again. And you still don't use the new technology, which has a special understanding that you can show it where it is in the room. So this is why we say we need manuals done by AIDAR servers because it's easy, cheap and very very fast to do it. And then we say if you have something, you have like big fluctuation or like a new of a lot of new people hiring, that it makes sense to invest the money and make a digital twin and then do the training. the virtual reality, or you are producer of the machine and you have like a lot of customers and your machine is not changing every year, that it makes sense to make this kind of stuff, this kind of training for sure. machine is a product and you sell that again and again and again, in this case, you create that one time virtual reality model and you can use it and use it, use it again. Well, where you're a custom machine builder and every of your machine is different, then AR skills would be the right thing as in a pretty easy way. You can set up things and keep it running and use it also to the customer side of things and create a value there. Correct, correct. And then, yeah, but if you have something like a process which is very dangerous, like explosion or high voltage or whatever, yeah, it makes sense to make it a virtual reality because then you can, you know, train the people, adjust the people, have their instincts trained, then they react very quickly if there's no time to react. because they already have this experience. And this is where virtual reality kicks in and you can do it very quickly. And I see that, you know, you see like very, very quick steps in generative AI. So for example, now I write a prompt and prompt is doing like a beautiful text or beautiful image. Two days ago, a meta was announced, something like prompt to 3D on the level, which is, looks very promising. So I guess the cost of 3D generating will be like going down. So you can have the assets for 3D trainings. You can also, this already in experiments which are doing like 3D models based on pictures, but they also understanding, you know, moving parts and this kind of stuff. So it was lowering the costs for that. actually answering already my next question because without asking me, because my question was actually, where do you see, does this whole technology go in immersive with AI and how does this all come together? And you already mentioned a couple of examples that you can create 3d models with a prompt or convert a 2D image into a 3D image with AI. Are there some other things you see out there as future trends? Yes, we are already experimenting with that. We call it AIDAR brain and this AIDAR brain is actually We connect, for example, what you see in AIDAR servers, the whole pictures, with we are streaming of our web RTC, like everything in front of you, your hand movements, and then the brain, so the artificial intelligence can do something with that. And what would be the examples? One is like, you can collect... pictures of errors if something is wrong in front of you and say, okay, label it and say, okay, this is a failure number one or this is whatever. And then this is building the grand truth, the data for argumentary or for artificial intelligence to build a model so you can do it. Another one is like, we are... we can send the movement of the hands to the back end and then say, okay, an expert was executing the action, and then the rookie will do the same. And we can compare the movement of the new guy with the expert, but not on the level that we have like a one -to -one, because somebody is more high, somebody is smaller, you are brighter, so the movements are different. Yeah, exactly. Yeah. So we have like a models which can compare to get the crucial parts. Like if you take it, take it not from above, but below, or you have to do it with the left hand or right. What is this kind of stuff? So this is where research development is going in the direction what we do. And then the next thing. umbrella of ergonomic analytics and stuff like that? Yes, also like, you know, like more make job more effective and also knowledge transfer because sometimes you don't even know what you do, that it is good or that's perfect. And if we can analyze the data and compare three experts, then we can say like, which kind of this operation is the most important stuff that it is better. Yeah. And the next one is like, you can just do your job and the artificial intelligence will create a tutorial based on it because it is understanding what you're doing, it's understanding the scene. So the editor would be very easy to use because you just do it. And then you see the whole timeline and you can just adjust it. So this is where I see the direction it is going to. And you see it's also like that. the time of content creation. So more or less, yeah, you show it one time and in the future AI will then create the instructions out of it and make it readily available. Maybe you have to tweak here and there a little bit, but the whole time you put into the content creation will shrink down. Exactly. amazing. Very good. Anything else on that area where you see potentials in the future in where this whole technology goes? Do you see on the IOT side any further improvements that you visualize current states, which is already possible as you describe? Are there any new things in that area? IoT is a good example because we are talking about IoT for years and it should be already there. The problem which we're facing now is there's no Wi -Fi connection on the production side where you can stream something because the Faraday effect is there. You know, we are fighting on the much lower level with problem that we already assume the technology is, yeah, like the adaptation of the technology. And the problem is like, you can have a beautiful wife on the production side, but the problem is like, okay, then the security somebody can kick in and do something wrong in the network. So it is better to, you know, keep your network secure with, you know, slow, whatever. Because nobody will say, okay, if, you know, like security is every time a good excuse for don't follow the technologies, just to say, okay, otherwise it would be like the problem. So, but I guess the people and we have to accept and we have to find a way and maybe artificial intelligence with something like automatically hacking of the system and giving like a proof that the secure maybe will, you know, reduce the problem or make it faster that the companies will accept to have nice internet. So if you connect Starlink as an example, if you know satellite internet, cheap one for if you compare to the cost, then you have like a high band. Usually if you build a new plant, it is like middle of nowhere, but with Starlink, you can still have that good internet connection. And then you can use all this process which you... your need. When this is the first challenge we are fighting with, you know, like this is the you have this would be like good future. And then, then I guess and it can be used. Yeah. But the problem is like, no more technology, less technology, but more stable and more like adapt by the people that we can do it inside of the company. a certain infrastructure. That's how I would define it out in the field to be able to use digital tools and the network and the wireless network, everything under the umbrella of wifi is definitely an important infrastructure. What I see out in the field is that this this infrastructure, the network and the wifi, it can be used for so many different things. More and more factories actually invest into that and build up that internal secure wifi network to enable all these other functionalities and features. So there's a change out there. But of course it always takes time to implement and to upgrade an infrastructure in an existing factory and adding more cables and responders and all of that. It's a costly thing that we have to keep in mind as well. Mm -hmm. Yeah. For sure. good. Let's dive a little bit into some use cases. Can you share some feedback from clients which have adopted the AiDAR Solutions solution? Maybe you can even tell some additional names and some specific use cases where clients used your product and what was the result there? Okay, I also already mentioned some of the examples, but one, for example, the Daimler was something like building a new plant and training people before it was there, without sending the people to another location that they can work as a shadow to another guy. This was a huge improvement. Another one is like, there was a company expanding very fast and they have like three plans over the world. And it was like the COVID time for them was like the remote support to have the possibility to send a checklist, something remotely and execute. This was very important scenario of advantage of our tool. And... Also like for companies which are building. machines and more like a one machine from one for one customer. And there's like a no big difference, but still differences for every customer. Then they have this, not the virtual reality, but the step -by -step tutorial. So again, create very easy, this kind of stuff and connect it to IoT and they have sessions. So these are the main advantages. And then we have, I'm not allowed to say the company names because they are very restricted to that, but. we know about that part. a lot of... They have one of part of those also robotic arms and they have to keep the training for 10 years. It means that in the training center you need like for 10 years all the equipment, all the younger and this is very costly also to have the experts. And now we are moving into virtual reality and to have like a standard procedure, virtual reality, how to train it. But with this, you know, one -to -one size. master memory or the stuff so they can do the training on digital twin of the machine and the people don't have to travel they can do it like remotely this is also which has to change in the mind of the people like if for training centers they use it the people are coming to them and they do it but they are changing and also the way how we execute the training. A lot of customers are saying, okay, then we have eight hours training in virtual reality and say, no, no, no, we don't. We don't do it the same way. It doesn't make sense. You know, in the virtual reality, you have like a hundred percent focus on what you're doing. You're not able to do all this. You, you know, check your phone or do something like that. You don't wait for the machine that you get the machine from you. This is exclusive for you. You have a trainer, which is exclusive for you. is like your private trainer, which is saying which step you have to execute. So you need less time to do the same. And one standard example is like we did like a virtual reality training for dentists and for 3D scanner for the teeth. And the normal... teaching was to invite doctors for two days and they have like a big presentations, PowerPoint slides, but then because the machines are expensive, it was only two of them and was 20 people in the room. They spent about 20, 30 minutes in front of the machine and do did something. After two days they have like an examination and everyone got their certificate. But after two weeks, nobody was memorizing anything. So it was like it was gone. And what we do is like we say, no, no, no, do like exclusive 20 minutes and then repeat it at 20 minutes. And then you and we can give you a feedback. So you have like a. We are training with assistant of us, so we show you what to do and then you go in the test mode and we don't show you. So you'll see, it is not already memorized. And after two days, again 20 minutes and then again, and then for sure it stay in your long time memory, then you can execute it. Yeah, the retention rate is way higher because you actually did it in a 3D environment. You used your hands and therefore the content goes into your mind versus you get that typical power presentation show. Yeah, it stays a while, but then it's gone. That is a pretty cool use case. I like that one. Mm. Very good. We talked a lot about all the different technologies under the immersive space and specifically how training in the future can be get better. Do you have some summarizing words at the end or something you missed to say about this whole content before we end this episode? You know, we like, we're still the producer of the product, so we say buy it. But to be honest, like, there is something like, you know, with this step -by -step and using the special understanding is something very easy to do and everyone should at least try to make and change the processes. And this is very easy step in the technology. where you can use it still as PDF, you can still use it on the tablet. But this is something if your majority level of your company is growing and adapting of the glasses, then you can do it. And I will say for everybody, start now, do it, and then grow with it, adapt it. And you see the technology is going in this direction, the artificial intelligence is helping with creating the content even very fast. So for me with Quest 3 and similar devices, this is the right moment to start because there is very convenient hardware on the very acceptable price level to go and introduce to your company. Yeah, I agree on that as well to start now and don't push it further and kind of wait. let's wait until this other new technology is further developed. Within digital transformation, you have to start and step by step going up and get yourself deeper into that. And as you mentioned, right now we are at a good point where the hardware is at a good price and then the good the software systems are already mature and developed and it's just a good way to start the whole thing. And how can people start? Of course, they can contact Marek directly, specifically when they are in Europe as Digitalitum is a partner of Aida Solutions. come to us, we're happy to do a demo to you and show it again, both solutions, the AR and the VR solutions, talk about your specific use case, the problems you wanna solve, and then find a solution how either solution can make or be the solution for that problem. Exactly, Marcus. Okay. Good. will put our contacts down in the show notes. Marek, if you have some videos you wanna share, we can put the links to that in the show notes as well. Yeah, so our listeners can get additional information and see a couple of videos on YouTube in how does actually look. Okay, perfect, no problem. All right, then thank you for being a guest at the Digitalitum Podcast. Thank you, Markus, for having me here. Okay, have a nice evening, day, whatever. Bye bye.