Skip to the content

S-Cube delivering breakthrough capability for waveform inversion of seismic data

Machine-driven earth-model optimisation. Industry's leading FWI algorithm. Uncover every detail.


XWI Parameter Learning

  • Fully automated data-in model-out optimisation framework for deriving the most detailed and accurate best-fit velocity volume from recorded seismic data. 
  • The toolbox integrates CWI and RWI with the AWITM and FWI cost functions for adjusting model training weights to predict seismic field recordings.
  • The result is more accuracy, more predictive power, more automation starting further away from the true answer.

Case Studies

A set of 3D field deployments comparing the velocity resolution achieved by waveform inversion relative to standard tomographic techniques.


Chevron NATS field data

Using XWITM to define and quantify the target layer formation in a producing field.


Equinor OBC field data

Using XWITM to capture details of a gas accumulation filled with fluid escaped from an underlying reservoir.


Woodside NATS field data

Using XWITM in exploration to reveal a prospect and pre-drill define its geometry. 


XWITM on AWS User Tullow Oil

Burst Compute Capacity

Industry's Leading Algorithm

Better. Faster. Deeper.  


Four pillars:



Predictive Power

Differentiated Capability

Cloud-native architecture:

Multiple XWITM initialisations

Travel through search space

Cloud-enabled hyper-tuning

Industry Awards and Measures of Esteem

Academic founders internationally recognized as leaders in the field

Distinguished Awards


Award for Adaptive Waveform Inversion

2015 Guido Bonarelli Award


Award for advances in Full Waveform Inversion

2013 Best Paper in Geophysics


Key Presentations


Woodside AWI deployment in Myanmar

Full-bandwidth adaptive waveform inversion at the reservoir (2017)


Statoil AWI deployment in the North Sea

Imaging Beneath a Gas Cloud in the North Sea without Conventional Tomography (2017)

S-Cube Cloud

Run XWI on the Cloud

Discover an unprecedented increase in the resolution of your velocity model.