Model-Based Development with MATLAB Simulink
Published on: August 10th, 2025 · Author: Rick Gijsberts
Model-based development uses graphical, mathematical, or textual models to describe software behavior; suitable models can also support simulation and code generation. GOcontroll Moduline controllers support model-based software development (MBSD) through MATLAB Simulink, with a documented Linux build workflow for the M1, L4, and HMI1.
What is model-based software development?
Model-based software development applies modeling to software logic and its verification. Model-based design provides the broader system-design context, including the process being controlled. In MATLAB Simulink, model-based programming expresses control logic through connected graphical blocks; not every model needs to generate code.
For wider embedded software development context, see embedded Linux control with modular hardware.
Model-based development: which targets are covered?
GOcontroll Moduline M1, GOcontroll Moduline L4, and GOcontroll Moduline HMI1 run Linux. They support programming with MATLAB Simulink, Node-RED, C, or the GOcontroll AI Application Builder.
GOcontroll Moduline S1 runs FreeRTOS and supports C, MATLAB Simulink, and the GOcontroll AI Application Builder. It has no Wi-Fi and must not be treated as a Linux target in the walkthrough below.
MATLAB Simulink toolchain and hardware blocks
The documented Linux workflow requires MATLAB Simulink, Embedded Coder, MATLAB Coder, Simulink Coder, Stateflow, and the GOcontroll Simulink Toolbox on MathWorks. The toolbox does not replace the required proprietary MathWorks add-ons.
The setup guide specifies R2023b or higher, but explicitly excludes R2025b from recommended use because toolbox installation does not work correctly. Use R2023b or another supported release. Install the toolbox through Add-On Explorer, using an installation path without spaces; the installation also includes required compilers.
Follow the setup and model-building instructions in the Moduline knowledge base. Create Linux-target models from the GOcontroll Linux template and build through Embedded Coder.
The toolbox exposes module I/O, CPU monitoring, voltage measurement, and LED control. Interface support must be checked for the target and available blocks. GOcontroll Moduline L4 provides onboard LIN master/slave hardware; using it from MATLAB Simulink depends on supported GOcontroll Toolbox blocks.
A model-based control design example on Moduline M1
An illustrative, untested closed-loop example could regulate water level using a voltage-output level sensor and a solenoid valve. Build the proposed configuration around the GOcontroll Moduline M1, a 6 Channel Input Module, and a 10 Channel Output Module.
| Configuration item | Documented parameter |
|---|---|
| GOcontroll Moduline M1 | 4 module slots |
| 6 Channel Input Module | 0–5/0–12/0–24 V inputs; frequency up to 10 kHz |
| 10 Channel Output Module | 1 A nominal, 1.8 A peak per channel; for solenoid valves and relays |
| Output-module sample time | 10–200 ms |
- Model the plant. Represent the proposed tank’s inflow, outflow, and level response separately from the controller. Define assumptions and acceptance criteria.
- Model the controller. Compare measured level with the target and define valve commands, including a deadband to limit repeated switching.
- Simulate test cases. Exercise setpoint changes, disturbances, and invalid sensor values. Record results against acceptance criteria; no measured results are claimed here.
- Map and build. Check sensor voltage, valve current, module channels, and target settings. Use the Linux template and build through Embedded Coder.
- Upload and test. M1, L4, and HMI1 support Simulink binary upload over Ethernet or optional Wi-Fi. Validate the deployed application against the defined tests.

GOcontroll Moduline modules must not be used by Simulink and Node-RED simultaneously. Each output module requires its own fused supply. Simulation does not replace testing on the configured machine.
Control logic can be reused across suitable Moduline configurations, but hardware mappings and target settings must be checked, and the application rebuilt and tested. For prototype planning, see the educational prototyping platform.
Where can this workflow be applied?
Documented agricultural uses include EOX E-OX 175 tractor functions—hitches, track width, steering, and dashboard—Equisport automated water management, weather stations, and climate control. Vehicle requirements belong in the special-vehicle application context.
Moduline controllers are also used in electrification systems. Predictive-maintenance functionality requires application-specific software and validation. Model reuse, simulation, and code generation are supported; development-time and reliability outcomes depend on the application and validation.
Frequently asked questions
What does MBSD mean?
MBSD stands for model-based software development. Models describe software behavior and support testing; suitable models can also support automatic code generation.
Which MathWorks products are required for the documented Moduline Linux workflow?
The workflow requires MATLAB Simulink, Embedded Coder, MATLAB Coder, Simulink Coder, and Stateflow, plus the GOcontroll Simulink Toolbox. Use a supported release; R2025b is not recommended because toolbox installation does not work correctly.
Do I need Wi-Fi to connect to a Moduline Linux controller?
No. GOcontroll Moduline M1, L4, and HMI1 support Ethernet for Simulink binary upload. Wi-Fi is available only with the optional hardware.
Key takeaways
- Model-based software development supports modeling, simulation, reuse, and code generation where supported.
- Use the complete toolchain and distinguish Linux targets from the FreeRTOS-based S1.
- Check hardware mappings, rebuild, and validate each selected configuration.
Latest blog
-

Interview with Huib from H2Consultancy
Selecting a fuel cell is only the beginning. The components, control system and application must then work together. Huib Hupkens van der Elst of…
-

Telematics Retrofit: CAN Data or Added Sensors?
Older machines can join connected fleets, but a CAN connection does not guarantee useful data. Audit available signals, sensor requirements, and platform access before…
-

Dutch Embedded Controller Suppliers: 3 Routes
Compare three sourcing routes: custom electronics, an embedded computer, or a modular machine controller. Define what the supplier delivers and what your OEM still…
