X-Twin
x is the unknown; in this field, it is the ground.
AI for engineering the physical world through one governed record of the asset, from planning and construction to operation and maintenance.
Readings, models, inspections and spatial context enter once, with identity, quality, time and provenance preserved.
Delivery asks how the asset is being built; Operations asks what it is doing now; Maintenance asks what is wrong and what should happen next.
Finite element modelling and physics-informed machine learning combine field evidence with BIM, GIS, asset relationships and engineering limits.
Developed at Monash University. X-Twin is led by Associate Professor Qian-Bing Zhang and built by his group and collaborators. Their research spans field sensing and geotechnical characterisation, multiphysics simulation and physics-informed machine learning, and the governed data engineering that keeps every prediction traceable to its evidence. The platform applies that work to underground and civil infrastructure, from site investigation to operation.
- ARC Future Fellow
- Editor-in-Chief, Tunnelling and Underground Space Technology
- Vice President, International Society for Rock Mechanics and Rock Engineering
- Director of Research, Department of Civil and Environmental Engineering, Monash University
Monash campus twin
A working digital twin of the Clayton campus: real buildings, ten monitored assets, 143 instruments and a tunnel drive passing beneath them. Select an asset, play the drive, move the view.
10 monitored assets
All instruments report through the same contracts a client's estate would use.
- Instruments
- 143 across 10 assets
- Types
- Settlement, tilt, crack, piezometer, vibration, weather
- Corridor
- 1.2 km of works
Campus geometry and building massing provide spatial context; for security purposes, only synthetic monitoring data is shown in this public demonstrator.
Operate the twin
Inspect assets, move through the drive, configure the spatial view and test an engineering review limit. Every control acts on the twin above.
- Asset ID
- MON-GRN-01
- Chainage
- 11.600 km
- Offset
- 0.0 m
- Instruments
- 104
Advance through the works
Choose an engineering view
Control what is visible
Test a settlement limit
No assets require review at the current model date.
The engineering method inside the twin
A digital twin is credible only when every prediction can be traced to evidence, assumptions and a versioned model. Explore the four stages of that chain.
Observe the physical system
Instrument readings, TBM telemetry, inspections, InSAR, BIM and GIS enter through governed adapters. Original payloads are retained; quality and provenance travel with every observation.
Measurements are admitted even when suspicious: quality flags preserve the reason for review without silently erasing evidence.
DeliveredX-Twin: one shared record, three applications
A digital twin platform for the infrastructure lifecycle. Readings, models and inspections enter one governed record; three applications read it for the phase they serve.
Ground model and alignment, construction parameters against the design envelope, settlement prediction by physics and AI, and the structures and utilities above the works.
Instrument readings and their freshness, condition against review limits, estate health and inspection capture. The campus twin above runs on this record.
Defect register over an asset-maintenance ontology, FMEA analysis, and repair prescriptions linked to their evidence and the relevant clauses.
Behind the applications: finite element modelling, physics-informed prediction, a 3D viewer, and admission of instrument data by supplier pull, direct push or field kit. Deployed on the client's projects with the client's identity provider, tenancy and retention rules.
Platform development, multiphysics modelling, sensing and consulting
Four lines of work, delivered by the research group with the platform as the working environment.
X-Twin deployed and configured for a client's estate, applications built on its published contracts, and integration with the client's own systems.
Enquire ›Finite element simulation of staged excavation, coupled ground and structure analysis, and physics-informed machine learning calibrated against the monitoring record.
Enquire ›Instrumentation schemes, gateway and edge integration, supplier acceptance testing, and admission of readings into the twin by pull, push or field kit.
Enquire ›Ground–structure interaction, settlement and damage assessment, interpretation of monitoring data, and independent review for tunnelling and underground works.
Enquire ›X-Twin Delivery
A recording of the application itself, unedited.
Contact
Enquiries about projects, asset estates, or software to run on the platform.
The applications run as a demonstrator. Request a login, or a guided demonstration of the platform on your own project.