Carnot transformed its data workflows from manual and fragmented to automated and reusable

The Data Management Challenge in Multi-Source Engine Development

Carnot is a UK-based engineering company developing ultra-efficient, low- to net-zero-emission engines for hard-to-decarbonise sectors including shipping, heavy-duty vehicles, and off-grid power. Engineering at Carnot spans a wide range of physical and virtual sources, physical test cells running National Instruments LabVIEW alongside simulation tools including Ansys Fluent (CFD), Ansys Forte and Converge (combustion CFD), Ansys thermo-mechanical, LS-Dyna multi-body, and Mathematica and GT-Suite for 1D modelling.

‍

The engine test data management challenge was not that data did not exist, it was that it was siloed. Post-processing ran through Python scripts with manual triggers, producing tables and images saved into folder structures with no context attached to the data at the point of visualisation. Comparing results between test configurations required opening separate files manually or running ad hoc processing. Combining datasets from different sources was done entirely by hand. Insights were limited in technical scope and in who had awareness of them.

‍

Automated Test Data Pipelines and a Unified Engineering Data Infrastructure

Carnot used Key Ward to build an automated engineering data infrastructure connecting their physical test cells and simulation tools within a single platform. Test data pipeline automation now ports data directly from cloud storage into structured, queryable datasets, the manual trigger step eliminated entirely. Automated KPI extraction surfaces engine performance metrics using Key Ward's native Python libraries and GUI interfaces, with processed data added to a centralized database where AI/ML regression methods can be deployed quickly without requiring specialist data science resources.

‍

Performance maps are generated from multiple operating points. Dashboards give engineers and stakeholders live visibility into the latest datasets, best historic results, and comparisons between test configurations with context attached to every dataset including test conditions, engineer responsible, and engine configuration. The manual errors and invisible context of the legacy approach have been replaced by a structured, auditable, and accessible simulation and test data management environment.

‍

What Is Coming Next

As the migration to Key Ward progresses, Carnot is exploring three additional capabilities. Key Ward's transformer nodes within pipelines enable time series analysis and anomaly detection,  surfacing insights from engine test data that would otherwise require significant manual investigation. A real-time connection between Key Ward and LabVIEW via JSON is in development, which would allow live monitoring of engine health and low-frequency data including temperatures, torque, and power during test runs. And by generating performance maps directly inside Key Ward, Carnot is evaluating whether the need for separate software tools such as UniPlot can be eliminated entirely.

‍

What the Full Case Study Documents

The full case study covers the pipeline architecture connecting Carnot's test cells and simulation tools to Key Ward, the automated KPI extraction and database workflow, and the dashboard structure enabling live comparison across configurations and against best historic data. It also documents the speed of migration from legacy systems and the roadmap toward a fully unified environment where test and simulation datasets are correlated and interrogated together. If your engineering organisation is managing test and simulation data across disconnected tools and manual post-processing workflows, the approach documented here is directly applicable.