How project developers, institutional investors, and lending banks use scanned assets in corporate VR environments and the role artificial intelligence plays
Real estate decisions are made at many tables: the developer’s, the investment committee’s, the credit committee’s. What all parties have in common is that they judge a physical object that appears in the process almost only indirectly: as a plan, a teaser, a photo set, an appraisal, or a data room. Each participant forms their own mental image from these sources, and the differences between those images often surface only once they have become expensive.
A toolchain that has matured considerably in recent years addresses exactly this point: mobile scanners capture the as-is state of an asset in hours, point clouds and imagery are turned into walkable 3D models, and corporate VR platforms make them a shared meeting place for distributed teams. That this is no niche topic is shown by a representative survey by Bitkom, Germany’s digital industry association: construction and planning is the top VR use case among German companies at 74 percent, and 57 percent attribute great importance to the technology for their own competitiveness. What the approach delivers in concrete terms depends on the vantage point.
Three perspectives on the same space
The project developer: align earlier, document continuously, sell better
For developers, the benefits begin before groundbreaking and do not end at handover. During planning and construction, clashes and misunderstandings between designers, trades, and owner representatives become visible in the walkable model while a change is still a mouse click rather than a change order. The rule of ten known from quality management describes why this shift forward pays off: every phase an undetected error survives multiplies the cost of fixing it.
Regular construction site scans additionally create a documented as-is state as of a given date. It serves as the basis for progress reports to buyers and lending banks, for evidence preservation and claims management, and it replaces part of the on-site appointments with committees, investors, and authorities. In sales and leasing, finally, the planned end state and the actual construction status can be shown in the same virtual space: prospects and acquisition teams walk the asset before it is completed, without traveling.
The institutional investor: broader asset access, standardized monitoring
At funds, insurers, and pension institutions, committees decide on assets that only a few of their members have ever entered. The scanned asset in a VR room changes this asymmetry: the entire investment committee gets the same access to the property as the two colleagues who flew out for the inspection, and discusses floor plate efficiency, layout, or alternative use potential on the object itself rather than on a photo set.
In the standing portfolio, the approach enables standardized digital walkthroughs across locations and countries. Asset and property management align leasing concepts and refurbishments on the model, and ESG and capex discussions, for instance, on decarbonization pathways, take place on the asset instead of in a spreadsheet. Scans taken at valuation dates, handovers, or for investor reporting provide objective evidence that can be archived and revisited years later.
The lending bank: walking the collateral, evidencing progress
For the bank, the asset is the collateral, and it is precisely this collateral that the credit committee can walk through in the future instead of leafing through it. In construction finance, drawdowns against progress can be underpinned by periodic scans as an as-planned versus as-built comparison. In the standing book, digital walkthroughs support ongoing monitoring, and in workout situations, all parties can see the condition of the collateral without everyone traveling. In syndications, participating banks gain asset access that used to fail on calendars and travel budgets.
Added to this is the governance value: the scan, together with the annotations from the walkthrough, can enter the credit file as an auditable record and documents the visual basis on which a decision was made. None of this replaces the physical inspection by appraisers and credit officers. It broadens asset access to everyone who would otherwise never make the trip.
Three tool layers and their concrete benefits
Corporate VR platforms and headsets: the shared place
Enterprise platforms such as Arthur or RAUM differ from consumer offerings through user and permission management, closed rooms, and defined hosting models. For financial institutions, this is the central point regarding banking secrecy, GDPR, and regulatory outsourcing requirements: where the data resides and who has access are hard selection criteria, and European or German hosting is a tangible advantage.
Standalone headsets of the Meta Quest class make entry inexpensive. The hardware is a modest one-off investment, while the running cost per session is low; conflating the two makes the format look artificially expensive. The real advantage lies in session quality: participants stand in the asset together, point, measure, and annotate at a 1:1 scale, with spatial audio and without the distraction of a second screen. The PwC study on immersive learning formats shows that participants in VR work with measurably greater focus and feel more emotionally connected to the content than in screen-based formats. Persistence comes on top: the room remains in place, annotations are retained, and walkthroughs are also possible asynchronously.
SLAM scanners, point clouds, and Gaussian splatting: capturing reality
Mobile SLAM scanners, handheld or body-worn, capture floors and entire buildings in hours rather than days, including on active construction sites and without elaborate tripod setups. The result is a point cloud: the measured geometry of what was built, not of what was planned. This is where the concrete benefits of this layer come from: a robust as-planned versus as-built comparison; reliable measurement and floor area calculation, especially in existing stock without current drawings; objective comparability through repeated scans, for instance, for construction progress or changes in condition; and evidence preservation as of a given date.
Gaussian splatting complements the measurement accuracy with photorealism: photos or video footage yield a faithful rendering of surfaces, materials, and light, valuable for marketing and condition assessment. Honesty about the state of the art is part of the picture: corporate VR platforms today process mesh formats such as GLB rather than native splat files, and standalone headsets impose tight polygon budgets. Between the raw scan and VR session, therefore, lies a preparation step. This is precisely where the third layer comes in.
Artificial intelligence: from raw scan to decision document
AI operates at four points along the chain. Before the room, it automates preparation: registration of multiple scans, denoising, remeshing the point cloud into a platform-ready model, polygon reduction, texture compression, and segmentation by floor and room. What used to mean weeks of manual work by specialized 3D service providers is increasingly becoming an automated pipeline, and with it falls the biggest cost barrier of the entire approach.
Second, AI extracts information from the scan: areas and room types; damage patterns such as cracks or moisture; comparison against the BIM design; and up to scan-to-BIM for existing stock. Third, it links the asset to the file: Document AI pulls lease terms, capex items, and appraisal findings from the data room and locates them on the model. Cash flow then hangs on the floor, the decarbonization pathway on the building, the defect as a marker on the building component.
Fourth, AI operates in the session itself and afterward: assistant functions answer questions about the asset and pull up documents, speech recognition records the walkthrough, and a language model turns it into minutes and a decision paper with located resolutions, including real-time translation in international syndicates. In an institutional environment, one cross-cutting condition applies: processing must take place under controlled conditions, for instance via EU hosting or locally operated models, since the data concerned relates to collateral, tenants, and exposures.
Limits and prerequisites
The approach replaces neither physical due diligence nor appraisal obligations, and it requires initial investment in a scanning and data pipeline as well as a confidentiality concept before the first scan is taken. Onboarding also needs to be planned: many participants are new to VR, which is why simple, robust session formats should come first and more ambitious scenarios later.
Conclusion
The division of labor can be put in one sentence: scanners capture reality, VR creates the shared place, and AI delivers the preparation, the information, and the record. For developers, this means earlier error detection, continuous documentation, and marketing without travel. For institutional investors, broader, standardized asset access across portfolios and borders. For lending banks, more robust and documented credit decisions across the entire life of a financing. The question is thus no longer whether the tools are mature, but which decision processes an organization moves into the room with first.
About the author
Christian Salow is managing director of altii GmbH in Frankfurt am Main. The B2B platform combines financial media distribution for institutional investors in the DACH region with audio and video production and sovereign AI infrastructure. Working at the intersection of financial communications and immersive media, he focuses on VR collaboration, spatial data visualization, and locally operated AI in institutional settings. He is co-initiator of the Immersive Finance Forum (IFF), which regularly brings finance professionals together in virtual reality.
* DE: Die ergänzenden Inhalte können KI-generiert sein. EN: The additional content may be AI-generated.