System modelling
Acausal, component-based models with ModelingToolkit.jl and Dyad for thermal, electrical, mechanical and fluid systems.
Home/Solutions/JuliaHub Applications
Julia · SciML · JuliaHub
We design and build applications on JuliaHub. We write physics-based and scientific machine learning models in Julia and wrap them in clean web interfaces for engineers, operators and customers. They are deployed on JuliaHub's cloud with compute that scales when needed.
JuliaHub application design
Valuable engineering models often live in one expert's scripts. We turn them into fast, tested Julia models with a usable interface, hosted on JuliaHub where the whole team can run them.
Acausal, component-based models with ModelingToolkit.jl and Dyad for thermal, electrical, mechanical and fluid systems.
Neural surrogates and universal differential equations trained on simulations, fast enough for real-time and edge use.
Interactive Genie.jl apps and Pluto notebooks with inputs, plots and reports designed for the people who use them.
Hosted apps with public or organization-only access, batch and GPU jobs for heavy runs, and compute that scales to zero when idle.
Calibrate and run models against sensor data from the ARP Device Cloud, turning a simulation into a running digital twin.
MATLAB, Python and spreadsheet models ported to Julia, with tests, version control and much faster solves.
How we deliver
Who uses the app, which inputs they change, and what answer they need from it.
Physics, data-driven or hybrid models, built as tested, versioned Julia packages.
Calibration with lab, test-bench or field data, and documented accuracy limits.
Genie.jl or Pluto front ends with sliders, scenario comparison, plots and exports.
Published on JuliaHub with access control, scalable compute and a clear path for updates.
Why Julia
Julia lets engineers write models that read like the equations and still run fast. JuliaHub adds the cloud workspace, compute and app hosting to share those models safely.
Example applications
How we engage
We review your existing equations, scripts or test data and agree what the app must answer.
A working Julia model and a first interface on JuliaHub, shared with your engineers for feedback.
Validation, tests, performance tuning and access control, ready for wider use.
Connect live data, add surrogates and new scenarios as the app proves its value.
JuliaHub and Dyad are trademarks of JuliaHub, Inc. The demo on this page is a simplified lumped thermal model solved in JavaScript for illustration. A production app on JuliaHub would run the equivalent Julia model.
Show us the equations or the spreadsheet. We will sketch the app and the JuliaHub deployment around it.