# SLAM scanner range: 70 m on paper, 50 m of good data – Eduard Chiru, FJD - Show: 2D Is Against My Religion (2DIAMR), host Michal Gula - Guests: Eduard Chiru, FJ Dynamics - Published: 23 Oct 2026 · 19 min · Expo & conference podcast · recorded LIVE - Topics: SLAM, LiDAR, Reality Capture, Spatial AI - Web page: https://www.2diamr.com/watch/slam-scanner-range-eduard-chiru-fjd/ - YouTube: https://www.youtube.com/watch?v=edZWw1MDRQ8 - Spotify: https://open.spotify.com/episode/7KAJjTVAE4zPCK2VN8bwXp - Apple Podcasts: https://podcasts.apple.com/us/podcast/slam-scanner-range-70-m-on-paper-50-m-of-good-data-fjd/id1878399775?i=1000797787675 ## Summary Recorded live at INTERGEO 2026, Eduard Chiru of FJD walks Michal Gula and Tomáš Barnáš through the X1 (tasks sent from the cloud, AI floor plans, voice notes translated by AI) and the T1, a SLAM scanner that also works as a TLS on a tripod – then gives honest numbers on usable range, speed limits for walking, car and drone mounts, and how FJD wants to stand out as SLAM scanners start to look the same. ## Key ideas - Per Eduard, the X1 is built for site managers: a task is created in the cloud and imported to the scanner, the operator sees where to walk, and the data goes back to the cloud for review, 2D plans and stockpile calculations. - Four of the five scanners FJD brought to INTERGEO can generate a 2D floor plan automatically with AI in FJD’s cloud platform; per Eduard and Tomáš, the walls come out well, while doors and windows still need to be added or adjusted by hand. - On the T1, switching to TLS mode on a tripod is an algorithm change, not a hardware one, Eduard explains: the scanner no longer moves, so it calculates points differently. One or two minutes per station gives the densest data; 30 seconds is enough for a quick pass. - Spec-sheet range is not usable range: per Eduard, the X1’s 70 m range gives about 50 m of good data day to day, and the T1 on a tripod holds its millimetre accuracy up to about 30 m. - Speed matters to the algorithm: per Eduard, about 1 m/s walking, 15–20 km/h on a car and about 5 m/s on a drone, with the mount mode selected before scanning; most SLAM hardware now looks the same, so FJD competes on software, algorithms and AI workflows in the cloud. ## Questions answered ### How much of a SLAM scanner’s range is actually usable? Less than the spec sheet says: per Eduard Chiru of FJD, the X1’s 70 m range gives about 50 m of good data in day-to-day use; beyond that the data gets noisy. (discussed at 11:51) ### Can a SLAM scanner work as a terrestrial laser scanner? The FJD T1 can, per Eduard Chiru: on a tripod in TLS mode it reaches about 4 mm at 10 m and 8 mm at 30 m, and the switch is an algorithm change, not a hardware one. (discussed at 8:22) ### How long should a SLAM scanner stay on a tripod in TLS mode? One or two minutes per station for the densest point cloud, per Eduard Chiru; about 30 seconds if you want to finish quickly and continue handheld. (discussed at 10:21) ### How fast can you move with a SLAM scanner? Per Eduard Chiru: about 1 m/s walking, 15–20 km/h on a car mount and about 5 m/s on a drone; above the limits the IMU or the algorithm can break the point cloud. (discussed at 13:25) ### Can SLAM scanners create floor plans automatically? Four of FJD’s five new scanners can, per Eduard Chiru: AI in FJD’s cloud platform draws the 2D floor plan, and doors or windows are adjusted by hand before saving. (discussed at 5:09) ## Quotes > “For example, the X1 has a range of 70 meters. So for example, I think in day-to-day use you can get maybe 50 meters of actual good data.” – Eduard Chiru (11:51) > “It's not a hardware adjustment, it's more like a algorithm adjustment.” – Eduard Chiru (8:59) > “We are not trying to only develop the hardware, because right now most of them may, may look the same. But what matters is the software, the algorithm.” – Eduard Chiru (16:11) ## Chapters - 0:00 A contact zoo at INTERGEO - 0:48 From a Bucharest reseller to FJD - 1:39 Six or seven shows a year - 2:40 Five new scanners at INTERGEO - 3:44 The X1: tasks sent from the cloud - 4:52 Automatic 2D floor plans with AI - 6:15 Voice notes translated by AI - 7:38 T1: SLAM and TLS in one device - 8:37 TLS mode is an algorithm switch - 10:13 One or two minutes per station - 11:05 How much of the range is usable? - 12:54 Drone, car and robo-dog mounts - 13:13 Speed limits from 1 m/s to 20 km/h - 14:15 Scanning from a scooter - 15:32 More SLAM companies every year - 17:36 AI beyond the chat window - 18:16 Wrap-up ## Transcript _Automatic transcription (ElevenLabs Scribe), lightly edited; times follow the audio._ **Michal Gula** (0:00): you just asked us, how many more, podcast episode we will be doing today. **Eduard Chiru** (0:05): Exactly. Yeah, yeah. **Michal Gula** (0:05): And I just realized that this is, like, a really cool job because you don't need to go all around those halls and, and from booth to booth. You can actually have people coming here. **Eduard Chiru** (0:15): Everybody come to see you. It's like zoo, right? Where like two animals sitting here-... and everybody's coming. they are not feeding us, unfortunately. **Michal Gula** (0:23): But, but we are that,.. how is, how is that zoo called where, where you can go and touch the animals? **Eduard Chiru** (0:29): Safari? **Michal Gula** (0:30): no, no. A contact zoo, yeah. **Eduard Chiru** (0:31): Con-contact zoo. I never heard of that. **Michal Gula** (0:32): Because you can actually come in and, and sit here and talk to us. **Eduard Chiru** (0:36): Okay. **Michal Gula** (0:36): Touch us. Feel free to take picture. **Eduard Chiru** (0:39): Yeah. I will. I will take some pictures. **Michal Gula** (0:42): Welcome. **Eduard Chiru** (0:43): Thank you. Thank you. Welcome. **Michal Gula** (0:44): FJD is a cool company. **Eduard Chiru** (0:46): Yeah, it's, it's pretty cool. We- **Michal Gula** (0:49): How long have, how long have you worked for FJ Dynamics? **Eduard Chiru** (0:52): Two years and half. Two years and half. **Michal Gula** (0:54): Okay. **Eduard Chiru** (0:54): Yeah, so I basically worked for a surveying company in Bucharest. And, that company were, was a reseller for FJD, and I got to play around a bit with the, with the SLAM scanners, and that's how I know about SLAM scanners in the first place, and then FJD. And later I did my Erasmus in Italy, in, in Bari, and, this company in Bucharest reached out to me and say, "Hey, FJD is looking for a guy willing to travel, willing to show our product." So I said, "Okay, why not? Let's try it." And since then it's, it's, yeah. **Michal Gula** (1:26): So now your, part of your job is to travel, yeah? Like, to be everywhere. **Eduard Chiru** (1:30): But almost 30%, yeah- **Michal Gula** (1:32): Okay **Eduard Chiru** (1:32): involved with traveling. So for example, after INTERGEO I will have a one month and half business trip, so that's quite a lot. **Michal Gula** (1:39): How, how many expos and conferences do you do during the year? **Eduard Chiru** (1:44): I would say six or seven. Or, of course I cannot attend all of them, but the local ones in Romania, this one, 3DISE, which I attended this year with, the company FORCAT- **Michal Gula** (1:54): Mm. Mm **Eduard Chiru** (1:54): if you remember. Yeah. **Michal Gula** (1:55): we didn't mention, but, agreeing to do this podcast, you also, like- **Eduard Chiru** (1:59): Submitted **Michal Gula** (2:00): committed to the next years of 3DISE conference. **Eduard Chiru** (2:02): Of course, of course. Of course, of course. Also with FORCAT, yeah. So I think for the 3DISE w- that, this was, this year was the, the second, let's say the second year. **Michal Gula** (2:13): Mm. **Eduard Chiru** (2:13): So I, I joined as well, and it's... I think it's a bit different than INTERGEO, so it's more, let's say, focused on, 3D Gaussian splatting. So 3D, this is your, let's say, religion. And INTERGEO, okay, you can see the product- **Michal Gula** (2:26): No quotes. Not, no quotes to be used. Like- **Eduard Chiru** (2:29): Yeah **Michal Gula** (2:29): 3D is against my religion, so 3D- **Eduard Chiru** (2:30): Of course. Yeah, yeah **Michal Gula** (2:31): forever. **Eduard Chiru** (2:33): Yeah, yeah. So, looking forward to it, yeah. **Michal Gula** (2:35): We did the interview in booth, yesterday. **Eduard Chiru** (2:38): Yeah, yeah, yeah. **Michal Gula** (2:40): what's, what's new that you brought to, the INTERGEO? **Eduard Chiru** (2:44): Yeah, so I think this year for, FJD, we, we have brought the, the most products so far. five, five of them. Let's say two of them are in the more entry level, position, so they're a bit cheaper. maybe three centimeters accuracy, so... But they are very easy to use, and everybody can, can use them and can get the data from them. So if, for example, if you are an interior designer or maybe an architect, you can use them to scan small apartments and so on. And then we have three that are more professional grade, scanners. For example, X1 or T1 that can be mounted also on the tripod. So those one are the, are the real game changers, for example. The, for the tripod one, we can reach up to four millimeters at 10 meters in, in terms of accuracy, so I think that's pretty, that's pretty good. Also this is, let's say, only the start with the, with the TLS combining with the SLAM scanners. In the future we'll have, of course, more products and better products. Yeah. This is only the beginning. **Michal Gula** (3:41): You brought us also one of them here. **Eduard Chiru** (3:43): Yeah. **Michal Gula** (3:43): Is it a flagship, or? **Eduard Chiru** (3:44): Not really, but it's also a very interesting scanner. This is the FJD Trion X1. what's new on this one is that it has an integrated, screen, basically. Also cameras for the 3D Gaussian splatting and all the information from the cameras, right? Visual SLAM and so on. Also with a small, GNSS receiver so we can get the, the georeference data. This is, two batteries included in, in the pack. Run time up to four or five hours. so this basically we, we produce this, this scanner mainly for the, t- the team leaders, site managers that have, let's say, more people to, to control. basically with this scanner you can just create a task on the, on the cloud. You can import the task in the scanner. The people who's, who does the actual scan on site can see where to walk and what, what sh- should he, he scan. After he finishes, he send it back to the cloud. The manager can, can review the data and can, do some, some basic workflows. For example, they can extract to different plans. They can do some, some bin modeling, stockpile calculations, and so on. **Michal Gula** (4:53): You told me yesterday interesting... Actually two interesting things that- **Eduard Chiru** (4:56): Yeah **Michal Gula** (4:57): I'm, I'm not sure whether it's, it fits for all the scanners, but you said that you can automatically generate the CAD drawing- from the data. **Eduard Chiru** (5:09): It's, almost for every scanner, yeah. From these five, four of them can, can do this, basically. Yeah. We, we try to implement this workflow in our cloud, creative portal, cloud platform, and basically when you, when you import the data there, the 2D floor plan can be generated automatically via some powered AI, AI functions. And of course, if something needs to be readjusted or, you need to remove the door or place it somewhere else, you can edit the, the 2D floor plan, and then finish it and, and save it as a final project. **Michal Gula** (5:41): for, for disclaimer, I think it's fair to say that, it's not like the full floor plan will be, like, ready to- **Tomáš Barnáš** (5:48): go. you need to- **Eduard Chiru** (5:50): Mm **Tomáš Barnáš** (5:50): like add windows or, or doors. **Eduard Chiru** (5:52): Yeah, yeah. **Tomáš Barnáš** (5:53): But, like, like the main things which is, adjusting the walls- set the walls and then everything else- like, this is working pretty well, right? **Eduard Chiru** (6:00): Exactly, exactly. You can manually add the, the windows and the doors, and then readjust the, the entire 2D floor plans, as the, the, as the reality, and then save it as a... They save or send it to your clients or, for a preview or, yeah. **Tomáš Barnáš** (6:16): And the second thing was, the audio annotation. **Eduard Chiru** (6:19): Yeah. That's, uh- **Tomáš Barnáš** (6:21): Audio annotation **Eduard Chiru** (6:21): also interesting feature. **Tomáš Barnáš** (6:22): Audio an- annotation, yeah. **Eduard Chiru** (6:23): Yeah. So, we, this is for the X1- **Tomáš Barnáš** (6:26): Okay **Eduard Chiru** (6:26): with the audio. You just plug in a, a microphone, and basically when, when the person who does the scan maybe sees a crack or a problem in a construction site, they can just take a picture for that, of that, problem. They can, record their voice, maybe this, the person who is scanning maybe is from other country than the site manager, so they speak different languages. But the operator of the scanner can speak in their own language. They can present the problem, and later in the cloud, the... let's say the AI will do the translation and the subtitles, and basically the site manager can hear in real time the, in his language, the problem mentioned by the operator of the scanner. **Tomáš Barnáš** (7:04): And it is also, like matching it with the image, which was done- **Eduard Chiru** (7:07): Of course **Tomáš Barnáš** (7:07): previously. **Eduard Chiru** (7:07): Yeah, yeah. And you can see it on, on 3D- **Tomáš Barnáš** (7:08): Correct **Eduard Chiru** (7:09): on the, on the point cloud. Yeah. So that's a, a new feature of, of the X1. **Tomáš Barnáš** (7:13): I didn't see that one. I mean, I mean, it's not the breakthrough from the point of view that you can just transcribe and translate the text. Like, that's not the cool part. The cool part is that you have already implemented in, in the workflow and, and it's also- **Eduard Chiru** (7:25): In the workflow, exactly, of the scanner. So basically you can see on, on 3D, on the point cloud exactly where the problem is, what needs to be done. Maybe the site manager can go on site actually to see if the problem needs to be solved or how it can be solved. Yeah. **Tomáš Barnáš** (7:38): Yeah, very smart. Yeah, yeah. can we speak also about the, I don't remember the name, but, l- like the top level, of the scanner. **Eduard Chiru** (7:47): It's the T1. It's the T1. FJD Trident T1. **Tomáš Barnáš** (7:49): Yeah. And, you showed me how it can work, like at the same time, like you can work with this as with the SLAM. And then if you decide, you can put it on a tripod- **Eduard Chiru** (8:00): Yeah **Tomáš Barnáš** (8:00): and it can work as a TLS. **Eduard Chiru** (8:02): Yeah, yeah. So let's say this is the most flagship product that we are releasing this year. It's called FJD Trident T1, and basically it's a combine- com- combination between terrestrial and mobile, scanners, right? So for example, if you are scanning something and you, for a specific area, you need that millimeter accuracy, you can just place the scanner on the tripod, and you can reach four millimeters, accuracy at, in 10 meters, right? So, and then for the rest of the scan, in the mobile mode, you can reach up to one centimeter. So if you only need one part that, that needs that higher precision, you can, you can do it. Hybrid mode. Yeah. **Tomáš Barnáš** (8:38): you said that when you want to use it as a TLS- you have to put it on tripod and then, change the mode. You need to tell the scanner that now you are scanning as a TLS. **Eduard Chiru** (8:48): Yeah, yeah. **Tomáš Barnáš** (8:49): does it mean it is doing some, like hardware adjustment? Or what's happening there if you, like need to tell it, like, "Now you are in TLS mode"? **Eduard Chiru** (8:59): It's not a hardware adjustment, it's more like a algorithm adjustment. Because- **Tomáš Barnáš** (9:04): Okay **Eduard Chiru** (9:04): basically the scanner will know, okay, I'm in terrestrial, so I need to calculate the, the points differently than in the handheld mode. **Tomáš Barnáš** (9:10): Okay. **Eduard Chiru** (9:11): Yeah. So this is basically the main difference between using it in, mobile mode or terrestrial mode. The hard- the algorithm will, will change a bit to, let's say, to accommodate that, that change. Because basically on the tripod, there's no movement at all. It's only the scanner that is rotating and capturing the points. So the algorithm needs to be a bit different to, to accommodate to that. **Tomáš Barnáš** (9:31): So maybe also i- if it's comparing the data taken in the TLS mode compared to the SLAM mode- they have like a priority, right? Those points. **Eduard Chiru** (9:41): Yeah, yeah. Of course. For example, if you are scanning, mobile mode, you finish with the mobile mode, you put the scanner on hold, let's say. You put it on tripod, and for that specific area, you repose the scanner. You do the... It takes around, I mean, depending on how much, density and accuracy you need, you can leave it 30 seconds, one minute, two minutes on the, to, to, to scan. And then basically you repose it, you take it out of the tripod, and start again to, to scan. This is the, the workflow, let's say. **Tomáš Barnáš** (10:14): How long would, would take, or how long should I leave it on a, on a TLS mode for the proper scan? **Eduard Chiru** (10:22): I think one or two minutes. Two minutes is maximum, so I think this is the higher quality that you can get. But if you are wants to, want to finish quickly, you can also leave it for 30 seconds- and then, continue with the handheld. **Tomáš Barnáš** (10:37): So really, because, it's constantly rotating, so you are leaving this longer only to get like a more dense point cloud. **Eduard Chiru** (10:46): Point cloud. Exactly, yeah. **Tomáš Barnáš** (10:47): Okay. **Eduard Chiru** (10:48): So the, the sensor, it has, 640,000 points per second. A range up to 120 meters. Of course, when the conditions are, let's say, perfect and, and good. Maybe some reflec- reflective objects, the, the points won't reach. But yeah, this is, uh- **Tomáš Barnáš** (11:06): this is an in- interesting thing. may- maybe you might not know the, the answer since you are not in a software dev- development side. But I'm always wondering because each of the companies we are speaking to, they always say like, "Okay, we have this range, like few hundred meters or what- whatever." But most of them are admitting that, "Okay, but we don't really use the data above like 50, 60, 70 meters- because they, they are not reliable enough. They have big noise and everything else." **Eduard Chiru** (11:34): Yeah. **Tomáš Barnáš** (11:34): So, is this something that you also like count with during the registration and alignment process so that you, like really get rid of the data, uh- like, like past some, some range? **Eduard Chiru** (11:47): Well, I would say yes for example, the, the X1 has a range of 70 meters. So for example, I think in day-to-day use you can get maybe 50 meters of actual good data, and in the, in the terrestrial mode for the T1 when you place it on the, on the tripod, the... basically the usable data, it's either, four millimeters in 10 meters or eight millimeters in 30 meters. So above that, the data is gonna be a bit noisy. You, you cannot count on it, let's say, 100%. **Tomáš Barnáš** (12:19): So on terrestrial mode you have, like, lower distance, shorter distance to have a good data or the opposite? **Eduard Chiru** (12:26): No. The... basically the data will be the same. The, the range- **Tomáš Barnáš** (12:29): Ah, okay **Eduard Chiru** (12:30): will be the same, but only the data that you can use- **Tomáš Barnáš** (12:32): Yeah **Eduard Chiru** (12:32): will be maximum up to 30 meters, right? And in 30 meters you have the, the eight millimeter, accuracy. If you go and you want to do some measurements above that 30 meters of where you place the, the station, the data can, can get a little bit noisy or cannot, let's say, sustain that millimeter accuracy range of, of accuracy. **Tomáš Barnáš** (12:54): Can you mount your dr- your, SLAMs on a drone? **Eduard Chiru** (12:58): Yes, we, we, we can. also this T1 can, can be mounted on the drone, on, on car. we also have the robo-dog mount. We also have an extension pole. Maybe you... if you want to measure some manholes or some rooftops you can, you can do that. Yeah. **Tomáš Barnáš** (13:13): So is there any, like, speed limit you wouldn't really would like to cross with the devices? **Eduard Chiru** (13:19): When you are, walking or, uh- **Tomáš Barnáš** (13:20): No, like when you have it on the drone or like on... you said, like it can be- **Eduard Chiru** (13:23): Yeah **Tomáš Barnáš** (13:23): on the car, on the robot, so. **Eduard Chiru** (13:24): We do have some- **Tomáš Barnáš** (13:25): Yeah **Eduard Chiru** (13:25): limits, let's say. it's not good to push above those limits because the IMU sensor or the algorithm may break the, the point cloud. But for example, when you're walking one meter per second, it's just normal walking speed. That's the, the most efficient way to scan. For the car mount, maybe 15, 20 kilometers per hour it's a, it's a sweet spot, and with the drone maybe five meter per second it's, also a good, uh- **Tomáš Barnáš** (13:51): O- once you, once you put it on the car mount, is there- **Eduard Chiru** (13:53): Yeah **Tomáš Barnáš** (13:53): any adjustment on the algorithm as well? Because you are telling- like another speed of, uh- **Eduard Chiru** (13:57): Of course. **Tomáš Barnáš** (13:58): Yeah. **Eduard Chiru** (13:58): Yeah. So basically with the, with the different, types of applications, you can s- before you start the scanning, you need to select I am in handheld mode, I'm in car mode, I'm in drone mode, and then the algorithm will, will adjust to that specific mode. **Tomáš Barnáš** (14:14): Okay. **Eduard Chiru** (14:15): Yeah. **Tomáš Barnáš** (14:16): do, do you need to set the mode only for the, like, real-time view alignment? Or like, for example, like I, I can, I can be walking with the device- and then I decide to hop on a scooter- during the same session. So I'm, I'm basically changing the mode, but I'm not telling the scanner I'm changing the mode. **Eduard Chiru** (14:43): Yeah, yeah. **Tomáš Barnáš** (14:44): So this can be reprocessed afterwards, or do I need to like stop scanning, change- -mode and then continue? **Eduard Chiru** (14:50): Well, the... in the best case scenario, yes, but you can still, let's say, outcome this. You can do the, the handheld, mode, and then you can jump on a scooter, on a bike. You can do the processing. The data will be fine, but, for example, depending on how much you s- you scan with the, with the handheld, how much you have scanned with the, with the scooter or the bicycle. Because if you put it on handheld mode, but you are on a scooter maybe eighty percent of the scan, then the algorithm will,.. maybe will, will have some, some errors. You really like scooter scanning, right? **Michal Gula** (15:26): This is, uh- I, I wanna get- Ah, **Eduard Chiru** (15:28): a good idea... **Michal Gula** (15:28): the job done as fast as possible, you know? **Eduard Chiru** (15:30): Yeah. Yeah, yeah. **Michal Gula** (15:32): Uh- No... you can see, like, most of the guests here notice that, every year there are more and more SLAM companies. what's your perspective on that? Like, okay, you are the ones of, of, like, established. th- this is far not your first year. It's- **Eduard Chiru** (15:50): Of course. So our first, SLAM scanner, we launched it in twenty twenty-two, so I think we already have four years of experience with algorithms, with sensors, with, integrated cameras, visual SLAM, and so on. But of course the, let's say, the SLAM domain, it's getting more and more crowded with, with different companies. But I think for ourself, we are not trying to only develop the hardware, because right now most of them may, may look the same. But what matters is the software, the algorithm, and basically the, the results that you can get and the, and the accuracy that you can get. We also... we are focusing on, on cloud, on cloud pro- platforms, where we have some automated, AI workflows. For example, we can extract automated 2D floor plans, as I mentioned. Or maybe if you are an interior designer, you can, do some AI rendering. I can push the, put the couch here. Oh, no, it does, doesn't look good. I, I put it here. Yeah, so some basic, modeling. If you are a forester, you can extract each tree individually with an ID, with crown width, with, all the dimensions, with carbon dioxide storage, and so on. And basically, we... what this, this lets you get the report and basically, lower your, your volume of work. So you don't have to do that much of a manual work on software trying to do every tree, one by one. With, with a cloud-based version, basically you can just input the scan, and with a few clicks you can get the reports that you can already send to your manager, to your boss, to your client. And, yeah, so this is our, let's say, vision for the... for FJD and for next years. We are trying to, to, to push more into this, let's say, AI workflows. Yeah. **Tomáš Barnáš** (17:37): Yeah, it's going to be heavy, the- **Eduard Chiru** (17:38): Yeah **Tomáš Barnáš** (17:38): the AI implementation. Yeah, yeah, yeah. Yeah, we, we have this discussion, now often because, like, AI has,.. like the boom of AI is, has passed- and now we see, like, real implementation- **Eduard Chiru** (17:49): Yeah... **Tomáš Barnáš** (17:49): and we are just asking everybody, like, what did you pick from the AI support- within, like, your software solutions? **Eduard Chiru** (17:55): Yeah. Of, of course the, let's say, the day-to-day person only knows the AI from, from the chat. Chat, yeah. They can just, yeah, chat. But here in the, let's say, in robotics, in sensors, in automation, this will be the, the real game changer, where the AI can basically transform the, the way you do your project or your job or, yeah, the, the scanning basically. Yeah, looking forward- **Michal Gula** (18:17): Thank you for coming **Eduard Chiru** (18:17): to follow. Yeah, no worries. Thank you. Thank you for having me. **Michal Gula** (18:20): good luck with, with your products. **Eduard Chiru** (18:22): Yeah. Thank you. **Michal Gula** (18:22): it looks like you, like, renewed the whole portfolio- **Eduard Chiru** (18:27): Yeah... **Michal Gula** (18:27): of the scanners. **Eduard Chiru** (18:28): Yeah, yeah. **Michal Gula** (18:29): So good luck with that, and thanks for coming, and hope to see you next year at- **Eduard Chiru** (18:32): Yeah, perfect... **Michal Gula** (18:32): 3DISE. **Eduard Chiru** (18:33): Yeah, of course. See you. See you. Thank you.