2D Is Against My Religion with Michal Gula

Smithsonian Starstruck is a 40-minute free-roam VR experience built from photogrammetry, satellite data, CAD files and astronomers' point clouds, then checked scene by scene by Smithsonian astrophysicists. Producer Tomáš Barnáš walks through the full experience, and director Elliott Mizroch explains how the project went from a Smithsonian brief to Venice Immersive 2026 in two years.What's covered:How they rebuilt real telescope sites in 1:1 scale from photogrammetry plus satellite dataWhy astronomers rejected the first night sky: that Milky Way is only visible from the other hemispherePillars of Creation built from Smithsonian Astrophysical Observatory point cloudsBlack hole visuals: the astrophysicist said "no fog, it's a disk", and nobody gets to step insideStandalone vs PC-streamed VR: Wi-Fi channels cap streaming at 16 headsets per roomHow a venue runs 100–120 headsets and around 1,000 visitors a day, and what a museum needs to host it00:00:00 How do you scan space? 00:01:33 What Smithsonian Starstruck is 00:02:52 MMT telescope from photogrammetry and satellite data 00:04:08 Getting the night sky scientifically right 00:09:40 Astronomer sign-off and exaggerating for VR displays 00:12:24 Big Bang and the galaxy room at relative scale 00:15:45 Galaxy interactions, Milky Way and the Sun's path 00:20:17 Hubble at 1:1 scale and Pillars of Creation point clouds 00:25:37 Director Elliott Mizroch on Smithsonian-grade science 00:30:45 Story vs science: the AR overlay compromise 00:34:08 Interstellar's black hole as the only reference 00:35:52 Two years, stars as the main character 00:38:53 How a Smithsonian VR project starts 00:42:13 Getting into Venice Immersive with a work-in-progress demo 00:46:07 Diamond planet, Chandra and Betelgeuse's supernova 00:53:00 James Webb at 1:1 scale and a black hole you can test 00:57:12 Giant Magellan Telescope and the ending 00:59:54 Venue tech: tracking patterns, ArUco markers, onboarding 01:04:00 Collisions and the VR field-of-view problem 01:06:00 Standalone vs PC-streamed VR: the 16-headset limit 01:10:33 Where to see Starstruck and Fever's business model 01:14:25 Traveling kits for events and museums 01:18:45 Wrap-upGuests: - Tomáš Barnáš, producer of Smithsonian Starstruck https://www.linkedin.com/in/tomasbarnas/ - Elliott Mizroch, director https://www.linkedin.com/in/elliottmizroch/ #VirtualReality #Photogrammetry #RealityCapture

NavVis CLX is a compact handheld SLAM laser scanner built for LOD200 reality capture, everyday site documentation, scan-to-BIM/CAD, quantity surveying and area referencing. Michael Schoelkopf, Director of Product Reality Capture at NavVis, explains how CLX compares with NavVis MLX and VLX, why dual LiDAR and an integrated screen matter, and where the scanner works—and where it does not.The conversation covers 400,000 points per second, three 12 MP cameras, live 2D/3D coverage feedback, control points, Cross Dataset Alignment, loop closures, IVION workflows, pricing, and Gaussian Splatting. Michael also explains when LOD200 is enough, why CLX is not the right tool for LOD300 or large feature-poor environments, and whether a non-surveyor can reliably scan a site.If you are considering buying a handheld SLAM scanners for construction documentation, scan-to-BIM, digital twins, floor plans, site assessment or high-frequency reality capture, this episode gives you the trade-offs, workflow logic and real limitations to evaluate before buying.Guest: Michael Schoelkopf — Director, Product Management Reality Capture, NavVis.00:00:00 Intro00:00:59 Why NavVis built CLX00:03:04 Cost, ROI and cheaper SLAM competitors00:07:00 What NavVis does: reality capture + IVION00:08:53 CLX specs, dual LiDAR and integrated display00:13:20 Who should buy CLX?00:16:13 Combining CLX, MLX and VLX data00:18:16 CLX vs MLX vs VLX: choosing the right scanner00:20:10 Best CLX jobs: LOD200, floor plans, quantity surveying00:23:13 Project scale and Cross Dataset Alignment00:25:44 Narrow rooms, minimum range and mirrors00:28:16 What NavVis sacrificed to make CLX smaller00:31:18 Where CLX fails: range, outdoors, tunnels and LOD30000:35:13 Camera quality and imaging trade-offs00:38:30 Gaussian Splatting beta and point-cloud measurement00:41:41 Live CLX scanning demo and on-device QA00:45:28 SLAM workflow: loop closures and control points00:46:52 CLX price, IVION and recurring costs00:47:44 Can non-surveyors use CLX?00:51:04 Where to test CLX00:54:36 Final advice: trust the data, not the product claims

INTERGEO, geospatial AI and reality capture are reshaping surveying, digital twins, BIM and the economics of technical trade shows. Olaf Freier joins Michael to explain how GPS, GIS, drones, SLAM and AI compressed decades of innovation into year-by-year disruption—and why face-to-face trust matters more as digital information becomes easier to generate and harder to verify.They break down the exhibitor mistakes that destroy trade show ROI: treating attendance as the strategy, choosing booth space without considering traffic, leaving directory profiles empty, and arriving without measurable goals for leads, brand awareness, product visibility or thought leadership. Olaf also explains why events remain powerful for comparing scanners, support ecosystems and real workflows—not just specifications.The conversation covers Olaf’s move from satellite surveying to organizing INTERGEO, the rise of AI-powered reality capture, scanning restrictions around critical infrastructure, show-floor stories, networking, startup visibility, workflow automation, the Application Dome, the Reality Capturing Hub, and the hands-on experience park planned for Munich.For geospatial companies, surveyors, reality-capture teams and exhibitors, this episode is a practical guide to choosing technology, building trust and getting measurable value from major industry gatherings.00:00:00 Intro00:00:43 Warm-up00:03:05 From survey engineer to INTERGEO organizer00:08:02 Early exhibitor lessons and event logistics00:15:25 GPS, GIS and AI: three industry shifts00:20:44 Drones, SLAM and faster innovation cycles00:23:20 The biggest exhibitor mistakes00:29:51 Setting goals for trade show ROI00:34:30 Startups, market leaders and shared KPIs00:37:39 How small booths win attention00:42:24 Organizer-exhibitor collaboration00:44:18 Why trade shows matter in the AI era00:48:26 New leads, market research and trust00:52:03 INTERGEO stories and show-floor chaos00:55:45 Scanning rules and critical infrastructure00:58:39 Why visitors should attend INTERGEO01:02:50 Comparing scanners, brands and support01:07:52 Networking, digital twins and BIM01:10:49 From product demos to business outcomes01:14:42 Reality Capturing Hub and experience park01:17:29 Exhibitors, visitors and event scale01:20:12 Last-minute exhibitor packages01:22:22 Competitor placement and booth politics01:26:08 What excites Olaf about geospatial’s future01:29:17 3DISE partnership and final invitation

Autonomous 3D scanning, GPS-denied SLAM, and Level 4B robotics are reshaping surveying, mining, and reality capture. Exyn Technologies COO Ben Williams explains the difference between automation and autonomy, why autonomous LiDAR systems outperform rigid workflows in dynamic environments, and what happens to scanning operators as machines take over field capture.The conversation covers autonomy levels, ExynAI and Nexys, underground mine mapping, digital twins, SLAM drift, GNSS and RTK corrections, point clouds versus voxels, Boston Dynamics Spot, autonomous drones, survey accuracy, robotics economics, OEM partnerships, and the shift from one operator per scanner to managing fleets of data-capture systems.Ben also breaks down when autonomous systems are economically justified, why dangerous and inaccessible environments lead adoption, how IPO scrutiny can improve enterprise credibility, and why the future of reality capture may expand the market rather than eliminate surveying careers.00:00:00 Automation vs. autonomy00:02:01 When automation should remain automated00:04:06 Risk, variability, and the autonomy decision00:07:31 The five levels of autonomy00:11:55 Drone autonomy and Level 4B00:17:52 Is “autonomous” becoming meaningless?00:19:18 Why uncertainty demands autonomous systems00:21:22 Safety, efficiency, and adoption00:24:19 The economic breakeven versus human labor00:29:56 Will autonomous scanning replace surveyors?00:34:46 Why reality capture is not a zero-sum market00:38:23 Users invent the best scanning use cases00:42:05 Ben Williams: engineering, Navy, and deep tech00:47:01 Joining Exyn and taking the company public00:53:24 ExynAI, Nexys, drones, robots, and vehicle mounts00:57:38 Autonomous underground mine mapping01:00:58 Point clouds, voxels, and unexplored space01:05:06 SLAM drift in GPS-denied environments01:10:48 Accuracy versus speed: SLAM and TLS01:12:46 Boston Dynamics Spot: demo versus field reality01:15:17 How robots choose routes and inspection paths01:18:00 SDKs, APIs, OEMs, and integrations01:20:00 Partners, pricing, and business model01:24:18 Continuous digitization and the 3–5 year roadmap01:26:33 Hardware versus software strategy01:29:47 Final partnership call

Digital twin, reality capture, BIM, point clouds, 3D mesh, LiDAR, SLAM and Matterport all collide in this sharp debate with Kelly Watt. We break down what actually makes a digital twin useful: not the buzzword, but the data, context, decision support, workflow and ROI behind it.Kelly explains why a 3D model, point cloud or Matterport tour can be part of a digital twin ecosystem without automatically becoming a mature digital twin. The discussion moves from iPhone LiDAR and Polycam scanning to TLS accuracy, SLAM drift, survey control, construction adoption, autonomous scanning, DFW airport, facility management ROI, in-house reality capture teams and the future of AI infrastructure.This episode is for reality capture professionals, laser scanning companies, BIM managers, facility managers, surveyors, construction tech teams and digital twin strategists who want a clearer way to separate marketing language from real operational value.00:00:00 Digital twin debate starts: is a 3D mesh enough?00:01:37 BIM, point clouds and 3D models inside digital twins00:03:48 Schrodinger’s digital twin and the missing data layer00:05:35 Why vendors call everything a digital twin00:06:30 iPhone LiDAR, Apple marketing and reality capture language00:09:07 Polycam vs TLS vs SLAM: democratization or replacement?00:10:05 Accuracy, drift, scale and use-case limits00:12:58 When a business problem does not need 3D00:14:21 ROI, stakeholders and digital workflows00:15:27 DFW airport and the 80/20 rule for digital twin programs00:17:17 Digital twin as a process, not a point in time00:18:08 Minimum viable digital twin: data synergy and decision support00:20:35 Time-based scans, real-time data and maturity levels00:21:31 Matterport, live cameras and rudimentary twins00:22:43 Should Matterport scans be called digital twins?00:31:20 Kelly Watt’s reality capture background00:35:32 From reality capture to digital twin programs00:46:39 Continuous scanning and change management00:49:42 TLS, SLAM, Matterport and 360 cameras in industrial plants00:52:06 In-house scanning teams vs service companies00:59:33 Why construction is still hard to digitize01:05:49 Future of reality capture and autonomous scanning01:07:55 Why beginners must understand TLS before trusting SLAM01:14:11 Will robots replace scan operators?01:15:24 Autonomous scanning, airports and operational ROI01:32:19 Motional digital twins, data centers and AI infrastructure01:43:17 Closing thoughts