Home/Solutions/JuliaHub Applications

Julia · SciML · JuliaHub

Engineering models, shipped as apps.

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.

JuliaJuliaHubModelingToolkit.jlDifferentialEquations.jlSciMLDyadGenie.jlPluto.jl
Motor thermal model · what-if appRuns in your browser
Winding temperature (°C)Insulation limit 155 °C
Peak temp-- °C
Margin--
Steady state--
Solve time-- ms

JuliaHub application design

From equations to an app your team opens in a browser.

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.

System modelling

Acausal, component-based models with ModelingToolkit.jl and Dyad for thermal, electrical, mechanical and fluid systems.

Scientific ML surrogates

Neural surrogates and universal differential equations trained on simulations, fast enough for real-time and edge use.

App and UX design

Interactive Genie.jl apps and Pluto notebooks with inputs, plots and reports designed for the people who use them.

Deployment on JuliaHub

Hosted apps with public or organization-only access, batch and GPU jobs for heavy runs, and compute that scales to zero when idle.

Models fed by live data

Calibrate and run models against sensor data from the ARP Device Cloud, turning a simulation into a running digital twin.

Migration and speed-up

MATLAB, Python and spreadsheet models ported to Julia, with tests, version control and much faster solves.

How we deliver

Rigorous model, simple app.

01 · Scope

Decision first

Who uses the app, which inputs they change, and what answer they need from it.

02 · Model

Julia and SciML

Physics, data-driven or hybrid models, built as tested, versioned Julia packages.

03 · Validate

Against reality

Calibration with lab, test-bench or field data, and documented accuracy limits.

04 · App

Interface

Genie.jl or Pluto front ends with sliders, scenario comparison, plots and exports.

05 · Deploy

On JuliaHub

Published on JuliaHub with access control, scalable compute and a clear path for updates.

Why Julia

Fast like C, readable like Python.

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.

  • One language from prototype to production model
  • State-of-the-art differential equation solvers and SciML
  • Apps your engineers can update without a separate web team
  • Heavy parameter sweeps on cloud CPUs and GPUs

Example applications

  • Motor and drive sizingThermal and electrical what-ifs for duty cycles, like the demo on this page.
  • Battery and thermal managementPack temperature, ageing and cooling strategies across operating scenarios.
  • HVAC and building energyPhysics-based building models calibrated with live sensor data for control and savings.
  • Process and plant modelsTanks, heat exchangers and reactors for operator training and optimization.
  • Hybrid predictive maintenancePhysics plus data to estimate remaining useful life from IIoT signals.

How we engage

Start from the model you already trust.

1

Model review

We review your existing equations, scripts or test data and agree what the app must answer.

2

Prototype app

A working Julia model and a first interface on JuliaHub, shared with your engineers for feedback.

3

Harden

Validation, tests, performance tuning and access control, ready for wider use.

4

Evolve

Connect live data, add surrogates and new scenarios as the app proves its value.

Have a model stuck in a script?

Show us the equations or the spreadsheet. We will sketch the app and the JuliaHub deployment around it.