Treedis ×

Three DFW facilities, live on one operational platform

A review of the DFW × Treedis connected worker pilot — and what the next twelve months look like.

ONE VIEW ECUP Central utility plant SEAVAULT Southeast Lighting Vault PIO PIO Building
DFW International Airport  ×  TreedisSeptember 2026

Matterport captures it. Treedis puts it to work.

A strategic Matterport partnership unlocks the power of the digital twin — Matterport builds the photorealistic capture, Treedis turns it into the place where daily operations run.

Built on the leading twin platform

Treedis is a certified Matterport partner — every DFW space starts as a Matterport capture.

From picture to workplace

On its own, a twin shows the space. Treedis layers work orders, live data, procedures and AI on top of it.

Proven at enterprise scale

The same stack already runs with some of the world's largest organisations.

Trusted by teams at
Apple Amazon Granite Construction Anheuser-Busch BP
The same capture with the Treedis working layer Treedis
Live data
Ram hydraulic
158 bar in range
Work order
WO-1187 · Press brake service
Open
The working layerTreedis pins work orders, live data and procedures onto the same space
The empty Matterport capture Matterport
The digital twinThe empty capture — photorealistic, but still only a picture
The partnership2

What Treedis puts inside a digital twin

Maintenance

Work orders, asset history and procedures, opened on the asset itself.

Knowledge

Manuals, submittals and training material, pinned where the equipment stands.

Operations

Live sensor readings and threshold alerts on the 3D tag.

Engineering

Drawings and P&IDs tied to physical locations, relationships mapped.

Safety

Permits, hazards and checks anchored to the space they govern.

Augmented Reality

All of it visible on site, overlaid onto the physical environment.

Facility digital twin cutaway
Work order
WO-4821 · AHU-2 belt replacement
Chiller plant · Assigned J. Rivera
OpenFrom Octave Attune
Workflow
SOP · Pump isolation
Step 3 of 9 — close discharge valve
In progress3D + AR guided
Live data
Supply air temp
54.2°F in range
Pressure 12.4 psi · via Tridium BAS
Engineering
P&ID M-201 · Chilled water loop
Linked from Autodesk drawings
Linked to asset
The platform3

TAI — the layer that reads the building for you

One AI layer trained on your twin, your documents and your procedures — six ways to put it to work, from building the twin to answering the technician standing in front of the pump.

TAI Chat

Talks to your team

Ask where an asset is and it walks you there in 3D. Describe a fault and it drafts the work order — on desktop or on a phone in the field, in the worker's own language.

TAI Knowledge

Captures your know-how

Drop in a document, or a video of an expert doing the job, and TAI turns it into step-by-step work instructions anchored to the exact asset in 3D — before that knowledge walks out the door.

TAI Tagging

Builds the twin itself

Point it at a 3D scan and it recognises the equipment and links it to your asset registry automatically — matching "Pump P-101" to "Centrifugal pump 101" on its own. New spaces arrive already tagged.

TAI Decision

Anticipates what's next

TAI watches your operations and, when something's wrong, recommends the next action with the evidence behind it: the alert, the trend, the history, the procedure.

TAI Vision

Sees what you see

Show TAI a photo of a leak, a gauge, a worn part. It reads the real-world state, grounds it to the right asset in the twin, and tells you what it likely means.

TAI Vision reading a photo and grounding it to an asset
On-site AR

Grounded in the real world

The same intelligence, on the floor. Through the AR app a technician looks at the actual pump and TAI knows which one it is, what's open on it, and what the procedure says to do next.

The platform4

What we set out to prove

DFW runs 26 square miles and 18+ landside buildings. The pilot took three of them to prove the model for all of it: one 3D interface that ends the hunt across the campus and across the applications.

01

End the search

New technicians lose hours locating equipment they've never been sent to before.

02

Break the silos

Work orders in one system, asset data in another, telemetry in a third.

03

Prove it in three buildings

ECUP, the Southeast Lighting Vault, and the PIO Building.

04

Connect the systems of record

Octave Attune for CMMS, Tridium BAS over an MQTT feed.

05

Put procedures in the space

PDF SOPs rebuilt as 3D step-by-step guides, tested in AR.

The pilot5

Plan versus reality

What we committed to
What went live
Unify spatial context, maintenance records and telemetry in one interface
Every asset carries its documentation from Autodesk, its work orders and its live operations data in the same 3D spot, pulled live
Prove the concept at ECUP and the Southeast Lighting Vault
Proved at ECUP, Seavault and PIO — 3 sites, 684 assets tagged
Read-only work order viewing from Octave Attune
Live CMMS integration — work orders and asset history across all three sites
Stream BAS telemetry into the twin via MQTT
Live MQTT feed — 3,000 sensors streaming onto the 3D tags
Convert PDF procedures into 3D SOPs and test AR navigation
3D SOPs built and AR tested at Seavault, plus step-by-step navigation routes at all three sites, on desktop and in AR
Not in the plan: TAI chat, querying site data within and across all three locations
The pilot6

Where the pilot stands today

3
facilities live — ECUP, Seavault, PIO
684
assets tagged, with documents, work orders and live data in one place
3,000
live sensors streaming into the twin
2,378
work orders pulled in live from Octave Attune — and counting
20+
SOPs rebuilt in 3D and AR across the three locations
The pilot7

The one goal we haven't closed

The pilot proved the technical foundation — the integrations hold, the data is live, the technicians can use it. What it did not produce is baseline operational numbers, because pilot usage wasn't instrumented for time-on-task. Measuring that is the first phase of the next twelve months, not an afterthought to it.

MetricHow we baseline itFirst read
Time to locate an asset Timed walkthroughs at Seavault and ECUP — guided in the twin versus unguided, same tasks, same technicians Q1
First-time fix rate Octave Attune work order history, before and after, on the asset classes covered by the pilot Q2
Workflow compliance SOP completion and step-skip logs captured in the platform Q2
Onboarding speed Time for a new technician to reach independent dispatch in a pilot facility Q2
Platform and integration uptime Continuous monitoring of CMMS and BAS sync consistency Ongoing
📌 Omer — internal, remove before send Can Robert help pull first ROI figures before the meeting? If yes, part of this table becomes results, not plan. Also sanity-check the metric list.
The pilot8

The next twelve months

Three tracks running in parallel: more buildings, more depth on the assets we already have, and the intelligence layer that only becomes possible once both are in place.

Footprint

  • Scale to 300 users across facilities and maintenance teams
  • Three buildings out to the 18+ landside buildings and the wider campus
  • Full DFW GIS map and Matterport inventory brought into one interface
  • Every new space arrives tagged by TAI, not by hand

Depth

  • Truly bi-directional CMMS — today work orders sync in; next they're raised and updated in Treedis too, and open in the field through the app as it happens
  • Operations alerts set in the platform — thresholds on the live feed that flag the right team before a fault escalates
  • Engineering drawings linked to assets, so every diagram has a physical location on desktop, mobile and AR
  • Asset relationships mapped — what a valve feeds, what goes down with it, before anyone touches it
  • Scale SOP coverage — generated from DFW's existing videos and documents
  • Field AR hardware trials, including Meta Ray-Ban

Intelligence

  • TAI advisor — recommendations drawn from work orders, telemetry, drawings and SOPs together
  • Faster reactive response: what failed, what it affects, what fixed it last time
  • Condition-based maintenance — sensor trends triggering work before the failure, not after
📌 Omer — internal, remove before send Framing call: pitch this as a 12-month track, a 3-year track, or show all 3 years individually? Slide currently shows 12 months.
What's next9

Where this ends up

A living twin of the entire airport — every building, every system, every technician working from the same picture.

The whole campus, one interface

18+ landside buildings, the GIS map and the Matterport inventory in a single front door — no more knowing which app holds which answer.

Systems of record, unified

Octave Attune, Tridium BAS, Autodesk drawings and SOPs behind one 3D surface. The integrations proven in three buildings, at airport scale.

A workforce that never hunts

AR routes to any asset, procedures in the space, onboarding in days instead of weeks — institutional knowledge that stays when people retire.

TAI running through all of it

New spaces tag themselves, any question gets answered from DFW's own record, and the twin starts recommending: what to check, what to fix, what's about to fail.

The pilot proved the foundation. The next step is the three-year enterprise agreement that takes it campus-wide.

IntelligenceTAI tagging, chat and advisor over everything below
Live operationsTelemetry, alerts and work in motion, streaming in real time
Systems of recordCMMS, BAS, engineering drawings, SOPs — connected, not replaced
The campus twinEvery DFW building, mapped once, kept current
The vision10