A February 2026 seminar brought together three organizations focused on spatial data and infrastructure technology to address a central industry question: what constitutes a digital twin, and when does it deliver operational value?
Moderated by Blaine Horner of Merrick & Company, the session featured presentations from Esri's Andrew Carey on digital twin architecture, gNext Labs' Russ Ellis on AI-powered defect detection, and Langan Engineering's Russell Hall and Brock Saylor on indoor mapping workflows. A consistent theme emerged: converting raw spatial data into actionable intelligence remains the industry's most challenging problem.
Esri — ArcGIS Reality Engine | Andrew Carey, Sr. Business Development
Redefining the Digital Twin
Andrew Carey challenged attendees to distinguish between digital representations and genuine digital twins. A rough sketch, SLAM-scanned model, or archived LiDAR dataset represents the physical world but does not constitute a functional digital twin in an operational sense.
A true digital twin functions as a connected system capable of answering specific operational questions: Where are assets located? What quantities exist? What outcomes result from specific scenarios? How will conditions change? These capabilities require integrating IoT sensors, live camera feeds, utility network data, and 3D spatial context into a unified platform accessible beyond engineering specialists.
The System of Systems Framework
Esri approaches digital twin architecture through three integrated components:
- A system of record where data resides consistently and can be reliably retrieved
- A system of insight providing analytical tools that extract understanding from the data
- A system of engagement making that data accessible to stakeholders who need it
This framework addresses a common failure mode: organizations that invested in LiDAR scanning years earlier cannot locate or access the resulting data today.
Case Studies: Nottingham and Raleigh
The City of Nottingham, UK deployed a digital twin to resolve a development permitting bottleneck limiting investment. By consolidating aerial surveys, ground control, and building scans into one accessible platform with public-facing interfaces for development feedback, the city fundamentally restructured its permitting process. The return on investment reached two pounds per pound spent.
The City of Raleigh, North Carolina partnered with Nvidia and Microsoft to merge 3D aerial scan data with static CCTV feeds and AI video analytics. Using Nvidia's VSS chipset, the system identifies stalled vehicles, counts pedestrians, and analyzes traffic patterns in real time — automatically triggering dispatch through the city's traffic management system. Interoperability through common location standards transformed this from parallel tools into an integrated digital twin.
Oriented Imagery and Web-Performant Delivery
Oriented imagery workflows within ArcGIS link raw source images directly into processed photogrammetric scenes, surfacing the highest-fidelity available image of any location with a single click. These images typically exceed processed output quality and prove especially valuable for inspection and documentation. The ArcGIS Reality Engine advances with photometric rendering, true ortho processing that eliminates building lean and geometric distortion, and web-performant delivery pipelines that make complex 3D models browser-accessible.
gNext Labs — AI-Enabled Defect Detection | Russ Ellis, President
The Inspection Bottleneck
Russ Ellis, President of gNext Labs, articulated the core problem: traditional infrastructure inspections produce 144-page PDFs that stakeholders rarely review. His platform replaces static deliverables with live, collaborative, AI-annotated systems. Every detected defect receives a unique identifier, GIS location, quantified measurements, and trackable history — enabling comparison across inspection cycles and predictive condition modeling. The system directly addresses how to allocate limited maintenance budgets most effectively.
Bridge Inspection
Ellis demonstrated a 3D model of a Scottsdale, Arizona bridge with toggleable defect layers — cracks, spalls, patches — each individually geolocated and quantified. The system classifies patches as defects, enabling inspectors to evaluate whether prior repairs persist or deteriorate over time. Outputs integrate directly into CAD, GIS layer files, and CSV/Excel formats for existing workflows.
Airport Runway and FAA Vegetation Clearance
The runway surface inspection demo visualized crack and patch relationships from altitude to ground level, supporting budget allocation decisions across full pavement surfaces. The FAA vegetation clearance module maps growth against mandated approach angle thresholds, color-codes severity areas, and generates reports with precise GPS locations, parcel boundaries, and property ownership data — providing information necessary to initiate contact and remediation.
Communication Tower Inventory
GeneXT's tower inspection module uses AI to parse drone imagery and construct structured equipment databases: height, equipment type, manufacturer, and antenna azimuth for every component. The platform supports ad hoc design work within the same environment, enabling pre-work modeling of new equipment placement or antenna reorientation.
Langan Engineering — Indoor Mapping with Mobile LiDAR | Russell Hall & Brock Saylor
SLAM Scanning and Scanner Selection
Langan Engineering's indoor mapping practice centers on two Leica SLAM systems: the Arc Backpack offering true 360-degree imagery and the BLK2GO providing approximately 270-degree coverage with compact flexibility. SLAM — Simultaneous Localization and Mapping — builds precise point clouds as operators walk through spaces without requiring static setup or GPS. Hall emphasized that vendor inventory and support response time represent the most important scanner selection variables. Leica's support infrastructure has proven available at 1:00 AM during difficult access projects.
Leica Pinpoint Registration
Hall demonstrated Langan's registration workflow using Leica Pinpoint, which imports multiple scans, colorizes them, and provides cross-section views for alignment. Scans align sequentially, snapping together when proximity thresholds are met, with overlap statistics and error readouts confirming quality. The registered cloud flows into TopoDot for extraction, AutoCAD for Indoor GML formatting, and various viewers for delivery. Langan employs four viewer platforms depending on client preference: Esri, TrueView, Bentley iTwin Orbit, and TopoShare.
Regulatory Context and K–12 Demand
Approximately 20 states now require indoor mapping for K–12 school facilities, with some explicitly mandating GIS-based delivery. California, Texas, Arizona, Utah, and Colorado rank among the most advanced. Typical mandates cover accurate floor plans and documented life-safety equipment locations. Saylor described a recurring pattern where critical institutional knowledge exists only with individual employees. At Chapman University, a retiring facilities manager held essentially all campus infrastructure knowledge in memory. The project converted that knowledge into a GIS-based system of record accessible to the full facilities team.
Platform Architecture and Application Tiers
Saylor described three application layers within indoor GIS ecosystems:
- A status monitoring dashboard for facilities managers and leadership
- A general explorer application for public or campus-wide navigation
- A mobile update application enabling field maintenance staff to update asset records without desktop GIS skills
A live demonstration showcased the near-complete City and County of Broomfield, Colorado deployment, featuring full building floor plans, oriented imagery at every scan position, and clickable room views tied to BLK2GO captures throughout facilities.
The Bigger Picture: What This Session Tells Us
The digital twin definition debate carries strategic rather than semantic importance. Organizations investing in 3D modeling and spatial data infrastructure achieve fundamentally different returns depending on whether they build digital representations or connected systems answering operational questions. Practitioners who begin from customer problems and work backward to data architecture construct more valuable systems.
AI is shifting infrastructure inspection from static documentation to condition prediction. When every defect carries a unique identifier, GIS coordinate, and measurement history across multiple inspection cycles, the dataset becomes a condition trajectory rather than a snapshot. GeneXT's platform exemplifies what predictive maintenance resembles when underlying data infrastructure supports it.
Indoor mapping is transitioning from specialized service to regulatory requirement. The K–12 legislative mandates Langan described represent a leading indicator. As requirements expand into healthcare, higher education, and corporate campuses, organizations with established indoor mapping workflows will absorb substantial demand.
Data accessibility represents the final mile challenge. All three presenters returned to identical challenges: collection and processing represent solved problems. Making outputs accessible, maintainable, and useful to non-technical stakeholders — and keeping them current over time — constitutes the remaining work.
Interoperability serves as the prerequisite for all else. Every effective system presented built interoperability as a design principle: common location standards, native CAD and GIS workflow integration, and delivery pipelines designed for browsers rather than engineering workstations.
