- End-to-end ownership
- Data people act on
- Enterprise-ready integration
QDT-IoTLAB
IoT Implementation
Connect sensors, gateways and enterprise systems into one operational picture.
View detailsTelecom
Physical AI Telecom Twins
Interactive 3D twins of cell sites and in-building networks, with a copilot that finds the fault and drafts the work order.
Overview
Operable digital twins of telecom sites that combine 3D geometry, radio coverage and live telemetry with an AI copilot for diagnosis and dispatch.
Physical AI Telecom Twins model a site as it is built (the mast, antennas, radios, cabinets, floors, cabling and the equipment around them) and overlay radio coverage and 15-minute performance counters on that geometry. Robots and drones in the twin inspect the site the way a field crew would: a drone orbits a sector, a rover drives to the cabinet, a quadruped walks an office floor with a scanner. A copilot combines their findings with telemetry to name the likely faulty asset, then drafts a prioritised work order with crew, parts and service-level target. The two demos below run in the browser on synthetic sites and data; production deployments are scoped to an operator’s own sites, inventory and monitoring feeds.
The demos use synthetic sites, telemetry and faults, and a built-in copilot rather than a live language model. Integration with an operator’s inventory, alarms and performance data is defined during scoping.
Use cases
- Outdoor RAN cell-site operations
- In-building 5G / DAS coverage and faults
- Remote site inspection with drones and robots
- NOC fault triage and field dispatch
Live demos
Open a twin in your browser.
Both run on synthetic sites and data. Best on a desktop or laptop; the first load downloads the 3D site models.
- OutdoorRAN cell-site twinA 24 m monopole with three n78 sectors, radio units and a ground cabinet. Sector B is running hot; a drone and a rover help the copilot find out why.Drone and rover first-person viewsSector lobes and 3GPP coverage mapInject a cabinet door faultAsk the copilot, then create a work orderOpen the live twin: RAN cell-site twin
- IndoorIn-building 5G DAS twinA three-level office with a DAS head-end, fibre to remote antenna units and a leaky feeder in the car park. A quadruped surveys the floors to trace a coverage hole and interference.Floor-by-floor RSRP and SINR mapsWalk-test with a quadruped and scannerCoverage-hole and PIM runbooksSection cut through the buildingOpen the live twin: In-building 5G DAS twin
Features
What it does
Capabilities are configured per deployment and agreed during scoping.
- 013D twin of an outdoor cell site: monopole, sector antennas, radio units, cabinet and compound
- 023D twin of an indoor 5G distributed antenna system (DAS) across office floors, head-end and parking
- 03RF coverage heatmaps based on 3GPP propagation models (TR 38.901 UMi and InH)
- 0415-minute RSRP, SINR and VSWR trends per sector or probe
- 05Drone, rover and quadruped agents with first-person camera views
- 06Fault injection and diagnosis, such as a cabinet door alarm, a fibre fault or PIM
- 07Copilot that highlights the suspect asset and explains the evidence
- 08Work orders with severity, SLA, suggested crew and parts, exportable as JSON
Benefits
What it means for your teams.
- Operations teams see a fault on the asset it affects, not only as a counter on a dashboard.
- Fewer wasted site visits: the twin narrows the likely cause before a crew is sent.
- Field crews arrive with the right parts and a work order that explains the evidence.
In practice
What Physical AI Telecom Twins looks like day to day
An illustrative walkthrough of typical use, based on the product's features. Not a customer case study.
Illustrative scenario
Finding a hot radio with Physical AI Telecom Twins
A network operations team triaging a degraded 5G sector (illustrative, synthetic data).
09:10
Alarm on the twin
Sector B shows high VSWR and a hot radio unit, highlighted on the mast rather than only in a counter list.
How it works
From telemetry to work order
How the twin links network data, the physical site and field operations.
From telemetry to work order
How the twin links network data, the physical site and field operations.
Press Play the flow, or hover and click any step to see what it does.
Our approach
Step by step, with you.
Each stage ends with something you can review, so decisions stay visible and reversible.
- 01
Model
Build a 3D twin of the site from drawings, inventory or a phone or drone capture.
- 02
Connect
Attach radio coverage, alarms and 15-minute performance counters to the assets they describe.
- 03
Diagnose
Robots, drones and a copilot inspect the twin and point at the likely faulty asset.
- 04
Dispatch
Raise a work order with severity, SLA, crew and parts, ready for your ticketing system.
What you get
Deliverables
- 3D twin of the agreed sites
- RF coverage and telemetry overlays
- Copilot runbooks for the agreed fault types
- Work-order export to your ticketing system
FAQ
Questions we hear
Does the copilot use a large language model?
The public demos use built-in diagnosis tools. A deployment can add a language or vision model, including one hosted on your own servers.
Can it model our own sites?
Yes. Sites can be modelled from drawings and inventory, or captured with a phone or drone, then linked to your alarms and counters.
Is the data in the demos real?
No. The demo sites, telemetry and faults are synthetic. A deployment uses your own sites and data feeds.
Which radio technologies are covered?
The demos model 5G NR n78 outdoor and indoor. Other bands and technologies are scoped per engagement.
Collateral
Collateral
Product brochure
Ask us for the latest material on Physical AI Telecom Twins, and a specialist will send it along with answers to your questions.
Start a conversation
See Physical AI Telecom Twins in your context.
Tell us about your sites, users and systems, and we will walk you through a configuration that fits.