Function-Call Subsystems in Simulink: The Most Overlooked Performance Booster

If you’ve been using Simulink for a while, you’ve probably built models full of Enabled, If-Action, or Atomic Subsystems.They’re familiar, flexible, and widely used , but not always the most efficient choice.

There’s another option that quietly outperforms them in many cases:Function-Call Subsystems.

Many modelers overlook them, assuming they’re only for library developers or those who work heavily with Stateflow. But when used effectively, Function-Call Subsystems can make your models faster, cleaner, and easier to maintain.

The Core Idea

If your model already uses function-call lines : signals triggered by events, thresholds, or Stateflow charts, you’re already in the right territory.

Instead of relying on Enabled or If-Action subsystems that execute conditionally but still evaluate every cycle, a Function-Call Subsystem runs only when explicitly triggered.

That means zero unnecessary computations and fully event-driven behavior, the way complex systems should operate.

Why They’re Worth Using

1. Precise Execution- The subsystem runs only when its associated event occurs. No idle processing in between.

2. Better Simulation Performance - By avoiding unnecessary updates, your model runs faster and scales better with complexity.

3. Cleaner Design - Each function-call subsystem can be designed, tested, and reused independently , making the entire model more modular and easier to debug.

A Practical Example

Imagine a vehicle engine temperature monitoring system with three subsystems:

• Cooling Fan Controller – activates above 90°C

• Warning Light – triggers above 95°C

• Data Logger – records data above 100°C

Without Function-Call Subsystems:All three subsystems are executed continuously, even when the engine is cool. This wastes simulation time and computation.

All subsystems are activated regardless of the temp.

Switches inside the normal subystems

With Function-Call Subsystems:A simple Stateflow chart monitors the temperature and triggers each subsystem only when its threshold is crossed.Each block runs exactly when needed — improving both speed and clarity.

Chart is used to generate function calls for the Subystems. (Apologies for the incomplete diagram, Matlab online does not offer stateflow license)

Light switch is on, only when the function is called.

The Takeaway

Function-Call Subsystems aren’t just for advanced users, they’re for anyone building event-driven or modular models.

If a part of your model only needs to run sometimes, it shouldn’t be evaluated all the time.That’s where Function-Call Subsystems shine.

Next time you find yourself designing logic that depends on specific events or thresholds, give them a try. You’ll end up with a smarter, more efficient, and easier-to-read Simulink model.