Introducing: EPCdoc’s Live Construction Monitoring Engine

Construction sites have always been the hardest part of an EPC project to see clearly. Engineering happens in an office — it is documented, reviewed, and filed. Procurement happens across a supply chain — it is tracked by PO number, delivery schedule, and inspection record. But construction happens in the physical world, in real time, across a site that might span hundreds of acres, with thousands of workers, dozens of subcontractors, and millions of individual activities happening simultaneously.

For decades, the only way to know what was happening on a construction site was to walk it. That is still how most projects are managed today.

EPCdoc LCM2 changes this. LCM2 — Live Construction Monitoring, second generation — is EPCdoc’s purpose-built AI engine for construction site visibility. It does not replace the site team. It gives the site team eyes they have never had before: continuous, objective, real-time awareness of everything happening across the entire construction site, processed by AI and delivered as actionable intelligence directly inside EPCdoc.

Why “Live” Is the Word That Matters

Most construction monitoring systems are not live. They are periodic. A drone flies once a week. A site engineer does a walkdown on Friday. Progress is reported at month-end. The word “monitoring” in their product names is aspirational — what they actually deliver is a slightly more automated version of the same periodic snapshot that has always existed.

LCM2 is genuinely live. The camera network processes footage continuously — not when someone reviews it, but in real time, every minute of every shift. The IoT sensor network transmits readings on a configurable interval — every 30 seconds for critical structural sensors, every five minutes for environmental monitoring. Drone data is processed and integrated into EPCdoc within two hours of a flight completing. The LCM2 dashboard reflects what is happening on the site right now, not what was happening last Friday.

This distinction matters because construction problems compound. A crane that has been idle for three hours this morning will not appear in Friday’s progress report — but it represents three hours of lost productivity on a critical-path activity that will push the completion date. A concrete pour that is running behind temperature specification today will produce a structural element that fails inspection in 28 days. A safety violation that happens at 7am on a Tuesday will not show up in the weekly safety audit conducted on Thursday.

LCM2 sees these things when they happen. Not when someone reports them.

---

The Five Core Capabilities of LCM2

1. Drone-Based Quantity Measurement

LCM2 connects to your site’s drone survey programme and does something no human can do from a drone flight: it compares the physical point cloud from the survey against the project’s 3D design model and computes actual installed quantities automatically.

Tonnes of structural steel erected. Cubic metres of concrete placed. Metres of underground services laid. Linear metres of piping installed. These are not estimated from a visual review of the drone footage. They are measured by the AI comparing what exists in three dimensions on the site against what should exist in the design model, at the resolution of the survey.

When the drone flies on Monday morning, the LCM2 quantity register updates by Monday afternoon. The progress figures in the EPCdoc project controls dashboard reflect measured installed quantities, not the site engineer’s Friday estimate. The client’s PMC representative, who has been asking for a defensible basis for your progress claim since month three, now has one.

2. Computer Vision Activity Monitoring

LCM2’s camera network does not record construction footage. It processes it. The distinction is important: a CCTV system gives you video to review after something goes wrong. LCM2’s computer vision gives you intelligence while the work is happening.

The AI models running on the LCM2 camera network identify and track:

Equipment utilisation — every crane, excavator, concrete pump, and piling rig in the camera field of view is identified, tracked, and its operating status recorded continuously. A crane that is rigged and positioned but not lifting is flagged as idle. LCM2 calculates equipment utilisation rates by shift, by day, and by work front — giving the plant manager the data to interrogate productivity without having to chase the subcontractor’s daily plant report.

Workforce density — the number of workers in each defined work area is counted and compared to the planned resource allocation. An area planned for a 50-person crew running at 28 people at 9am on a working day is a resource deployment problem. LCM2 sees it at 9am, not at the end-of-week labour return.

Activity classification — LCM2 distinguishes between different construction activities within each work area. Formwork being erected, rebar being fixed, and concrete being placed are three different activities in the same physical location that represent very different stages of completion. LCM2 tracks which activity is underway in each area, giving the construction manager a live view of work-front status without a site walkdown.

HSE compliance — LCM2’s AI models detect PPE non-compliance in real time: workers without hard hats, without high-visibility vests, working at height without visible fall protection. Non-compliance is flagged with a timestamp, a camera reference, and a location — not as a weekly audit finding, but as an immediate alert that goes to the HSE Supervisor and the Construction Manager simultaneously.

3. IoT Sensor Integration

LCM2 integrates with the physical sensor network that modern construction sites increasingly deploy — not as a data aggregator that stores readings for someone to review later, but as an active monitoring system that interprets sensor data against design specifications and raises alerts when readings fall outside permitted ranges.

Concrete maturity monitoring — sensors embedded in pours transmit internal temperature readings throughout the curing process. LCM2’s AI uses the temperature-maturity relationship for the specific concrete mix to predict when the pour will reach the specified minimum strength. The prediction updates as curing progresses. When a pour is ready to strip — not when the standard curing period has elapsed on the calendar, but when the actual concrete in that specific pour has actually reached strength — LCM2 tells the construction team. On a project with multiple pours per week, this optimisation alone recovers significant programme time.

Structural monitoring — sensors on temporary works, falsework, and early-age structural elements monitor deflection, vibration, and load. LCM2 raises an alert if any reading approaches a threshold that would indicate instability — before it becomes an incident, not after.

Environmental monitoring — wind speed, temperature, humidity, and precipitation readings across the site are processed against the work-method statements for activities in progress. When wind speed exceeds the crane operating limit in the area where a critical lift is scheduled, LCM2 alerts the lifting supervisor before the crane is mobilised to position, not after.

RFID and QR tag tracking — major deliverables (pressure vessels, structural modules, piping spools, equipment packages) are tagged in the fabrication yard and tracked through transport, site receipt, storage, and installation. LCM2 knows where every tagged item is at all times. When a tagged item that should have arrived on site last Tuesday has not moved from the fabrication yard, LCM2 flags the delay — automatically, to the procurement team and the planning engineer simultaneously — before it hits the critical path.

4. Clash Detection and Resolution Register

Physical clashes — where two installed systems conflict in space because site installation deviated from the coordinated 3D model — are among the most expensive construction problems to fix. Finding them before both systems are permanently installed saves weeks of rework. Finding them after is a variation order, a delay claim, and a damaged subcontractor relationship.

LCM2 performs continuous clash detection in two modes:

Pre-installation — before a new system is installed in an area, LCM2 compares the planned routing against the as-installed conditions in that area captured from the latest drone survey. If structural steel was adjusted on site and no longer matches the coordinated model, LCM2 flags the conflict to the piping or electrical team before they install into the area. This catches the category of clash that the 3D coordination review missed because site conditions had already changed by the time installation began.

Post-installation — LCM2 continuously analyses the drone point cloud for physical proximity violations between installed systems, measuring actual clearances against the minimum clearances specified in the design model. Proximity violations below the minimum clearance threshold are raised as clash records regardless of whether they were visible in the design coordination review.

Every clash record in LCM2 carries a severity rating, the two systems in conflict, the responsible engineering discipline, and a resolution workflow. The clash cannot be marked closed without a resolution record, an updated drawing reference, and a confirmation scan showing the clearance has been restored. Clash history is maintained against the project record — available for the as-built close-out and for any future maintenance or modification work on the asset.

5. AI-Powered Site Analysis and Prediction

The five data streams flowing into LCM2 — drone surveys, camera feeds, IoT sensors, tag tracking, and clash records — are individually useful. Together, they are the inputs to EPCdoc’s AI site analysis engine, which does three things no individual data stream can do alone:

Productivity forecasting — LCM2 combines installed quantity data from drone surveys with workforce and equipment utilisation data from cameras to calculate actual productivity rates by crew, by activity type, and by work front. These rates are compared against the planned productivity assumptions in the CPM schedule and projected forward to calculate realistic completion dates for each work package. When a work package is at risk of missing its planned completion date based on current productivity trend, LCM2 flags it — with enough lead time to do something about it, not after the milestone has been missed.

Critical path impact simulation — when LCM2 identifies a productivity problem or a delivery delay that will affect a work package’s completion date, it automatically runs the CPM network forward to identify which downstream activities are affected, what their new earliest-start dates are, and whether the project completion date has moved. This happens continuously, not once a month when the planning engineer updates the programme.

Anomaly detection — LCM2 learns the normal patterns of activity on your specific site: which work fronts are typically active at which times, what the normal equipment utilisation profile looks like for a productive shift, what the typical concrete pour completion rate is for your crews. When the pattern changes — a work front that is normally active goes quiet, equipment utilisation drops sharply in an area that should be at peak productivity, a pour rate that has been consistent for three weeks suddenly slows — LCM2 flags the anomaly for investigation. Not every anomaly is a problem. But every problem starts as an anomaly, and finding it when it starts is the difference between an early intervention and a delay claim.

---

LCM2 Inside EPCdoc — Not a Separate Tool

LCM2 is not a standalone monitoring platform that feeds data into EPCdoc. It is a module of EPCdoc — which means the data it generates is immediately available to every other module in the system.

When LCM2’s drone survey updates the installed quantity of structural steel, the EPM progress dashboard updates automatically. When LCM2 flags a tagged equipment delivery as delayed, the procurement module updates the delivery schedule and the CPM engine recalculates the impact on construction start dates. When LCM2 raises a clash record, it appears in the Document Controller’s transmittal queue as a request for a revised drawing. When LCM2 detects a safety non-compliance, it creates an HSE event record in the QMS module that triggers the investigation workflow.

This integration is what makes LCM2 a project management tool rather than a monitoring tool. Monitoring tells you what is happening. Project management connects what is happening to the decisions that need to be made about it — and makes sure those decisions reach the right people, with the right information, at the right time.

---

What LCM2 Delivers for Each Stakeholder

Construction Manager — a live dashboard showing work-front activity, equipment utilisation, workforce deployment, and any active safety alerts. The morning briefing is based on what actually happened overnight and this morning, not on what someone reported at the end of yesterday’s shift.

HSE Manager — real-time PPE compliance monitoring, environmental condition alerts, and a continuously-updated site safety status without waiting for a safety walkdown to happen.

Planning Engineer — productivity-based completion forecasts for every active work package, updated as each drone survey is processed, with automatic CPM impact calculation showing which activities are at risk.

Project Controls Engineer — measured quantity progress from drone surveys, replacing estimated progress from site walkdowns. The monthly progress report is built from data, not from negotiations with the site team about what the right number is.

Procurement Team — real-time visibility of tagged equipment and material location, with automatic alerts when a delivery is not tracking to schedule.

Client / PMC — a portal view of site progress backed by drone survey evidence, safety compliance metrics, and clash resolution status. The progress claim is no longer a number the client has to take on trust.

---

The Construction Site Has Never Been More Visible

LCM2 does not change how construction is done. It changes how clearly you can see what is being done — and how quickly you can act when what is being done is not what was planned.

The gap between what is reported and what is real on a construction site has always existed. LCM2 closes it — not by asking people to report more accurately, but by replacing the report with a measurement.