New release, new upgrades 🚀 Curious to hear from everyone, when it comes to product updates, what excites you more? Algorithm improvements or new software features?

Geometry changes everything. A year ago, we started building OpenGeometry around a problem that kept bothering us. We built a geometry kernel because we knew it was fundamental. Today, it’s clear. Everyone in AEC needs a solid geometry core. AI without control over geometry is just noise. Ionut is building his app using OpenGeometry and showed what’s possible. Today, we pulled together a quick end-to-end flow to see how far we could go. Text to floor plans. Floorplans to CAD. CAD to render. Under a minute! It all works. Not by chance, but because the geometry layer is strong. That’s the difference! You can do the same thing we did here. It’s open source. If you want the exact workflow we used, drop a comment, and I’ll share it. #cad #bim #aeco #opengeometry

🚀 Introducing LixelStudio 4.0: Engineering-Grade 3D Data, by Default For reality capture professionals, the biggest bottleneck isn't the scanning—it's the data cleanup. Tilted floors, patchy point clouds, and overly aggressive filters that erase crucial structural details can derail a project before drafting even begins. XGRIDS is changing that with the release of LixelStudio 4.0. Powered by next-generation core algorithms, it delivers perfectly leveled, photorealistic, and highly precise spatial data straight out of the box. 💡 5 Key Enhancements for High-Fidelity Digital Twins: 🎨 Photorealistic Coloring: Vibrant, true-to-site colorization that holds up even in challenging low-light environments like basements and parking structures. 📐 Auto-Leveling Precision: True horizontal and vertical alignment across massive datasets (including 24-story structures) with zero manual guesswork. 🔍 AI-Powered Filtering: Smarter artifact removal that strips out trailing noise while completely preserving thin structures, railings, and utility corridors. 🧊 Next-Gen Mesh Generation: Significantly fewer holes, crisper architectural contours, and sharper texture mapping closer to real life. 🖥️ Overhauled UI Workspace: Redesigned 3D clipping, slicing, and measurement tools built for a faster, frictionless workflow. Stop spending hours fixing data. Start drafting with engineering-grade precision from the moment processing finishes. 👇 Upgrade your reality capture workflow today: 📥 Download LixelStudio 4.0: https://bit.ly/48R03I2 #iTekMedia #XGRIDS #LixelStudio #RealityCapture #PointCloud #3DScanning #3DGS #GaussianSplatting #DigitalTwin #BIM #AEC #SpatialIntelligence #LiDAR #3DReconstruction

Your Digital Twin is a beautiful coffee table book. That’s the problem. We’ve all seen them. Those massive, gloss-finish Atlases that sit on coffee tables. They look expensive. They make you look like a "global thinker." But have you ever tried to actually find a specific village in the Pyrenees using one? Without a functioning index, you’re just flipping through 400 pages of pretty ink, hoping for a miracle. It’s an exercise in frustration disguised as "knowledge." In 2026, we are doing the exact same thing with Digital Twins. Companies are spending six figures on high-fidelity, LiDAR-scanned, 3D masterpieces of their facilities. They look incredible in a boardroom presentation. The CEO can "fly through" the vents. It’s the ultimate digital ornament. But when a pipe bursts in the basement at 3:00 AM, the facility manager discovers the "Atlas" has no index . The asset tags don’t match the physical valves. The maintenance logs are buried in a PDF that isn't linked to the model. The "search" function returns 404 errors because the data hierarchy was an afterthought. The result? Hours of wasted effort. The "Twin" sits there looking pretty while the building literally floods. A Digital Twin isn't a trophy; it's a spatial index. Its primary job isn't to look like a Pixar movie—it’s to reduce the time between "What is broken?" and "Where is it?" to zero. If you’re building a spatial model this year, spend 20% of your budget on the "pretty pictures" and 80% on the Index. A grainy, 2D floor plan with a perfect data index will beat a "broken" 3D masterpiece every single time. Stop building ornaments. Start building maps. #DigitalTwins #SpatialKnowledge #BIM #FacilityManagement #SmartBuildings #LighthouseInsights

Three types of data run every construction project. BIM models tell you what should be built. Point clouds tell you what is built. 360° images show you the texture and condition the geometry can't capture. Until recently, seeing all three together required three separate tools, two specialist workflows, and days of processing time. Watch what happens in a single browser tab. 1️⃣ A .LAZ point cloud of a bridge structure — loaded natively, full density, no downsampling. This is the as-built reality. Every beam, every surface, every deviation already captured. 2️⃣ One click to overlay the IFC design model. The as-designed and as-built sit in the same window. No export. No conversion. No waiting. 3️⃣ Split screen. Point cloud on one side, BIM on the other — same spatial coordinates, same moment in time. The eye catches what the numbers miss. This is what SpatialSense was built for — surveying firms and project teams at mid-market AEC companies generating LiDAR data daily, comparing it against design models, and writing deviation reports that nobody can open without the right software. Formats supported: .LAS, .LAZ, .PLY, .E57, .IFC, .GLB, .GeoTIFF, 360° JPG/ZIP. The AEC software market runs 86.8 tools per firm on average. None of them share data. $15.8B/year disappears into that gap. SpatialSense is one URL away from collapsing three of those tools into one. → spatialsense.app #ConstructionTech #BIM #LiDAR #PointCloud #AEC #SpatialIntelligence #DigitalTwin #ConTech #SiteInspection #BrowserNative #SpatialSense #Surveying

Have you tried Rhino + AI yet? It is a total game-changer for floor plan analysis. Watch our consultant Hue Fan Wong transforms 2D drawings into clean architectural data in seconds, catching scale errors and automating area calculations with ease. We are building a community of Rhino 8 power users—follow us to stay ahead with these workflows! Comment "AI" to join our Rhino 8 user group. #Rhino8 #ArchitectureAI #AEC #BIM #FORIDAAU

Our team have developed a new online tool for automatic #quality #assurance reporting for #precast #concrete components. In the instance below, we are identifying (surface) geometric deviations between the measured #point #cloud data and #BIM #design #model - DM if you'd like to try it out! Our next release will provide full internal reconstruction (e.g. of rebar dimensions, layouts, cover etc) using remote sensing paired with our new physical AI model for inversion

Automated geometric QA for precast — point cloud vs BIM, deviation maps, reports generated automatically. Built by Brian and the team. Next up: full internal 3D reconstruction with our physical AI inversion model. Then the same stack applied to in-service infrastructure. DM if you want to try it. InfraMind Labs

We use the Trimble X7 laser scanner system to scan multiple facilities, covering areas larger than 150,000 square feet. 3D scanning and modeling help bridge the gap between design and reality, saving project teams time and money. #EMCOR #BIM #DesignBuild #Prefabrication

321 million points. About 500 m × 750 m across, with 185 m of vertical extent. This video took 90 seconds to make. No meshing. No offline processing. No separate cinematic visualization workflow. Just PurpleGeo, fully interactive on a laptop. This is Niah Great Cave in Sarawak, Malaysia, from extraordinary laser scanning by Guy Van Rentergem. Caves are a perfect example of why navigation matters. Once you’re travelling through a long, narrow underground passage, it becomes incredibly hard to understand where you are in the wider structure. So throughout this video, we move back and forth between map view and 3D, keeping the detailed local view connected to the full scanned site. No RGB. No problem. Colour can be useful for recognising surfaces and adding context. But when you are trying to understand the shape of a cave, a tunnel, a rock face, or a mine, geometry comes first. RGB can even work against you: flat capture lighting hides structure, while baked-in shadows can make the surface look more or less pronounced than it really is. PurpleGeo’s realtime lighting responds to the actual scanned geometry, making the structure immediately legible. And when a dataset has taken this much skill and effort to capture, a 3D model you can simply tumble around is not enough. Tumbling makes intuitive sense for objects you can hold in your hands. It is a terrible way to understand hundreds of metres of terrain, tunnels, or cave. PurpleGeo lets you move through the space at human scale, stay oriented, understand where you are, and communicate it clearly. Huge thanks to Guy Van Rentergem for the remarkable dataset.