How MES Model & Refactor Improves Model-Based Design

Inconsistent Modeling Style

Imagine working on a Word document where everyone uses a different language. It would be impossible for team members to understand each other and collaborate. Well, the same happens when each modeler uses a different modeling style.

It is extremely important that people who develop models in Simulink or TargetLink adhere to common style guides that ensure models are understandable and maintainable for external users.

Several aspects need to be considered in model-based design, such as signal flow and how inports and outports are positioned within the model, signal line crossings which have to be kept to a minimum, or blocks and signal names which should be clearly visible and consistently positioned.

Error-Prone Models

Earlier, we noted that using different languages makes it very hard for team members to collaborate on the same document. A second factor worth examining is the structure. Imagine having the conclusion at the start, and the introduction in the middle, and so forth. It would be impossible to make sense of it all.

Something similar happens with poorly designed models. Failing to use a good subsystem structure for your models can lead to functional problems. A few best practices that should be followed are easy-to-understand separation of model parts, low complexity at the subsystem level, and separation of structural and functional levels.

Time Spent on Refactoring and Standardization

So, what does having models that are hard-to-understand and maintain mean for teams? Well, basically, team members will have to use a significant portion of their time on manual tasks such as layouting the models for a consistent style, changing the subsystem structure by moving blocks between levels, and renaming signals for consistency, among others.

This is a big problem because a large part of companies’ resources will be dedicated to fixing and standardizing models instead of actually working on the functionality. On average, this means that more than 30% of modeling time ends up being spent on refactoring and standardization tasks, such as laying out subsystems.

Reduced Productivity and Value Creation

Relying on manual efforts to adapt models, especially late in development, can greatly reduce a team’s efficiency. It is tedious, distracting, and makes changes costly. To top it all off, it also makes peer reviews inefficient due to constant layout discussions and the need to repeatedly fix models. This approach to model-based design is suboptimal.

The elements we mentioned before are only a few of the time-consuming tasks that can appear when modeling. Other tasks, like adding or removing cross-hierarchy signals or changing bus structures, can also take up a significant amount of time.