AI‑powered construction monitoring platform

They relied on periodic on-site inspections and spreadsheet reports. Drone footage existed for some locations but was never connected to project schedules, design data, or any consistent analysis. Progress updates to government bodies and investors were based on rough estimates rather than verifiable data.

S-PRO analyzed the client’s needs, reviewed available technologies, and identified Postdata as the right delivery partner. S-PRO then facilitated the introduction and kept the collaboration aligned with the client’s operations throughout.

Replacing Manual Inspections with Automated Monitoring

Each site required a physical visit, photos, and a manually assembled report. The process was slow, varied depending on who did the inspection, and didn’t scale across multiple projects. By the time a report reached a decision-maker, the data was typically one to two weeks old. The goal was to make drone flights the primary data source and have the platform produce progress assessments within minutes of upload.

Connecting Drone Data to Project Intelligence

Standard reporting was disconnected from design plans, timelines, and milestones. A project manager had no way to tell how far along a specific component was, or whether the project was on schedule. The platform introduced drone imagery to overlay against design plans, compare flights over time, and answer the question every stakeholder asks: are we on schedule?

Evidence-Based Reporting for Stakeholders

The client reports to government entities, investors, and contractors, all of whom need data-backed progress updates. Previously, these reports were assembled manually from scattered sources and depended heavily on individual judgment. The client needed structured reports that stakeholders could verify.

S-PRO’s role was to make sure the right problem was being solved by the right team. After assessing the client’s requirements, S-PRO coordinated with Etijah, who defined the tech stack and provided core expertise in geospatial data, computer vision, and construction technology. This collaboration led to the selection of Postdata, who built the platform through a structured engagement that moved from discovery to a production-ready product.

Construction AI: Automated Progress Detection

The AI engine analyzes drone orthomosaics against design plans and estimates completion at both the project level and for individual components. Each estimate includes a confidence score and specific evidence from the imagery: what was expected versus what was observed.

When a new flight is uploaded and processed, the AI generates a detailed report within minutes. Project managers see which components are progressing, which are behind, and where the current site diverges from the design plan. The next phase will add warnings for data quality issues – shadows, obstructions, or plan-versus-site mismatches – so teams know how much to trust each estimate.

2D/3D Visualization and Measurement

Drone data is processed into interactive orthomosaics and 3D point clouds viewable directly in the browser. Built-in tools handle distance, area, and earthwork volume measurements without external CAD or GIS software.

A comparison mode lets users view two drone flights side by side with synchronized navigation. CAD overlays allow design drawings to be placed on top of drone imagery with adjustable opacity, so teams can check whether construction matches the plans.

Flight Management and Data Processing

The platform handles the full drone data lifecycle: upload, automated processing through OpenDroneMap, and generation of orthophotos, point clouds, surface models, and reports. Teams upload raw flight imagery and receive processed outputs inside the platform – no manual file management or third-party processing tools.

Project Management and Collaboration

Work is organized into projects, phases, and milestones, with budget tracking, role-based access control, and status flags for on-track, ahead, and delayed projects. Annotations let team members pin observations to the 3D model or 2D map, flag issues, and assign follow-ups.

An executive dashboard gives leadership a portfolio-wide view: status distribution, progress trends, budget utilization, and a map of every active site.

Core components:

• AI progress analysis using Vision Language Models, with per-component confidence scoring
• Interactive 2D/3D viewer built on Potree, Three.js, and Leaflet with measurement and annotation tools
• Automated drone data pipeline with OpenDroneMap for orthomosaic and point cloud generation
• Role-based project management with phases, milestones, and budget tracking
• CAD overlay system for design-versus-reality comparison

Technology and infrastructure:

• Web-based platform accessible from any browser, no desktop software required
• OpenDroneMap for photogrammetric processing
• Vision Language Models for construction progress estimation
• Potree and Three.js for 3D point cloud rendering
• Leaflet with OpenStreetMap for 2D mapping and geospatial display
• Role-based access control with project-level permissions
• RTK/IMU/GPS telemetry support

Postdata’s delivery team:

• Product Manager / Business Analyst
• UX/UI Designer
• Software Architect
• AI/ML Engineer (VLM integration and prompt engineering)
• DevOps / Infrastructure

The client’s operations, project management, and executive stakeholders played an active role as product partners throughout – contributing domain knowledge, validating requirements, and reviewing builds. The collaboration ran primarily through offline feedback cycles rather than frequent calls, which kept reviews closer to the client’s actual workflows.

Where project managers previously drove out to sites and assembled reports by hand, they now have AI-generated progress reports within minutes of a drone flight. Each estimate includes specific observations from the imagery, confidence scores, and data quality warnings.

The 2D/3D visualization means anyone on the project can explore the site remotely, take measurements, compare progress over time, and flag issues directly on the model. Progress conversations with contractors, investors, and government stakeholders are now based on visual, verifiable data rather than written estimates.

The Construction AI module doesn’t produce a single completion percentage. It breaks a project down into its actual components. For a park development, the system analyzes the roundabout separately from the road network, separately from the park facilities, down to sub-components. You end up with a clear picture of which piece is running behind, not just a number.

Permissions were built in from the start. A contractor sees only the projects they’re working on; project managers and executives get the full portfolio view. That mattered here, where projects involve multiple stakeholders with different access needs.

S-PRO is leading the Phase 2 proposal and the conversations about expanding the platform to other government entities. The same approach – assess the client’s actual problem, find the right delivery partner, manage the engagement closely – is what they’re looking to repeat in other construction markets across the GCC.