I spoke with Povilas Sindriunas a Business Develoepr from  @UNDETPointCloudSoftware  to dissect one of the biggest hidden problems in the reality-capture industry.Everyone is obsessed with scanners, hardware, and field workflows. But the real bottleneck appears after the scan is finished.Point clouds are captured faster than ever thanks to SLAM scanners, mobile mapping, and cheaper hardware. Yet turning those point clouds into usable outputs — 2D drawings, BIM models, and engineering documentation — still takes the majority of project time.This conversation dives deep into the scan-to-BIM bottleneck, exploring how software like Undet attempts to reduce the time spent on:vectorizing point cloudscreating CAD drawingsbuilding BIM modelsquality control of outsourced modelinghandling massive datasets inside CAD environmentsThe discussion also explores the future of SLAM vs TLS scanning, automation in Revit workflows, the reality of outsourcing BIM modeling, and why full automation with AI is still far from replacing human modeling.If you work with laser scanning, photogrammetry, Scan-to-BIM, or point cloud processing, this episode explains where the industry is actually heading.Topics covered include:• Why scanning is no longer the bottleneck in reality capture• The real cost of turning point clouds into deliverables• SLAM vs terrestrial laser scanners in the as-built market• Automation tools for Revit modeling from point clouds• Handling massive point cloud datasets efficiently• Quality control of BIM models vs point cloud data• The reality of outsourcing scan-to-BIM work• Why AI is not replacing BIM modelers anytime soon• Future developments like SLAM data optimization and MEP extractionChapters00:00 – The real bottleneck in reality capture00:20 – Hardware vs software in scan-to-BIM workflows01:46 – Why point cloud vectorization takes most of the project time04:34 – SLAM vs terrestrial laser scanners: where the industry is going06:42 – iPhone scanning vs professional reality capture07:34 – Accuracy myths in as-built documentation10:29 – Is 2D CAD still relevant in the industry?12:09 – Povilas’ background in architecture and BIM software14:01 – Why join Undet and the future of digitizing buildings15:21 – Platform software vs plugin approach for point clouds20:40 – The main problem Undet tries to solve22:33 – Handling massive point clouds on weak workstations23:10 – Slicing point clouds into raster images for faster workflows29:00 – Supported point cloud formats and indexing workflow31:19 – Native Revit workflow vs Undet workflow34:11 – Automation tools for modeling walls from point clouds39:11 – Automatically placing windows and openings42:04 – Handling inconsistent measurements and model tolerances46:27 – Quality control: comparing BIM models to point clouds48:16 – Reality capture conference announcement49:11 – Undet software ecosystem explained51:45 – Why SketchUp users work with point clouds55:46 – ARES Commander vs AutoCAD for CAD workflows1:00:03 – Real projects and scan-to-BIM services1:02:20 – Outsourcing BIM modeling: quality vs expectations1:06:18 – Software vs services: why both exist in Undet1:07:55 – 3D Gaussian Splatting and why it’s not a priority yet1:09:24 – Upcoming SLAM data improvements1:12:29 – MEP extraction from point clouds1:14:07 – The reality of AI in Scan-to-BIM1:16:21 – Final thoughts and software trialKeywordsscan to BIM, point cloud processing, Undet software, Revit point cloud workflow, SLAM vs TLS scanners, reality capture industry, BIM modeling from point clouds, laser scanning workflow, CAD vectorization, scan to CAD automation.