Back office ops · Production

Rolls-Royce and Databricks optimize cGAN model training for aerospace design space exploration using Mosaic AI

The problem

Rolls-Royce needed to enhance design space exploration beyond the limitations of parametric models and traditional geometry modeling, while managing multi-objective constraints where requirements such as weight reduction and efficiency improvement conflicted with each other.

Workflow diagram · grounded in source
1
Design condition specified
trigger
“enabling the reuse of legacy simulation data to drive the identification and assessment of innovative design concepts that satisfy a specified design condition without requiring a traditional geometry modeling and simulation process”
2
Data table modeling
integration
“Data tables are set up to ensure they are optimized for the specific use case. This involves generating identity columns, setting table properties, and managing unique tuples.”
3
cGAN model training
ai_action
“the developed ML models are trained using a 2D representation of 3D results from typical simulation studies. This involves embedding knowledge of unsuccessful solutions to help the neural network avoid certain areas and find solutions fa…”
4
Design process implementation
output
“Once we developed and optimized models and algorithms, we would then implement them into the product design process”
5
Continuous model optimization
feedback_loop
“Based on current results, we plan to continually optimize the models and algorithms by adjusting parameters, refining the dataset, and ultimately changing the approach to handling multi-objective constraints.”
6
Secure model export
output
“The model trained with legacy data can be exported in a standard format, enabling the option of taking a copy to a secure environment where transfer learning can be conducted with project data characterized by a restrictive Export Contro…”
Reported outcome

The collaboration demonstrated faster time-to-model via AutoML and Managed MLflow, reduced total cost of ownership through a unified lakehouse architecture, improved model accuracy through scalable hyperparameter studies, and a governance framework through Unity Catalog for compliance-centric aerospace data.

Reported metrics
Total cost of ownershipsignificantly reducing costs
Time-to-modelfaster time-to-model
Model accuracyimproves accuracy
Model development and testing speedspeeds up the model development/testing process
Reported stack
Databricks Data Intelligence PlatformDatabricks Mosaic AIAutoMLManaged MLflowMLflowUnity CatalogRay
Source
https://www.databricks.com/blog/rolls-royce-mosaic-ai
Read source ↗

Frequently asked questions

What did this team achieve with this AI workflow?

The collaboration demonstrated faster time-to-model via AutoML and Managed MLflow, reduced total cost of ownership through a unified lakehouse architecture, improved model accuracy through scalable hyperparameter stud…

What tools did this team use?

Databricks Data Intelligence Platform, Databricks Mosaic AI, AutoML, Managed MLflow, MLflow, Unity Catalog, Ray.

What results were reported?

Total cost of ownership: significantly reducing costs; Time-to-model: faster time-to-model; Model accuracy: improves accuracy; Model development and testing speed: speeds up the model development/testing process (source-reported, not independently verified).

How is this back office ops AI workflow structured?

Design condition specified → Data table modeling → cGAN model training → Design process implementation → Continuous model optimization → Secure model export.

Related back office ops cases