Glossary · Automation software engineering and architecture
Automatic code generation
Also known as: Code generation, Generated code
German: Automatische Codegenerierung
In software engineering, automatic code generation is the production of source code by a tool from a higher-level input such as a model, a table, a template or a configuration, rather than writing the code by hand. In automation, code is generated for example from plant engineering data, state machines or Simulink-style models.
- Software engineering
In one sentence
Automatic code generation produces source code from models, tables or templates instead of writing it by hand.
Example
A generator reads the I/O list from the electrical CAD export and creates the PLC tag table and device function block instances for each conveyor section.
How it applies
- Engineering: Generation removes repetitive manual work and keeps code consistent with the engineering data it comes from. The source model or table becomes the Source of truth; generated code should not be edited by hand, or changes are lost at the next run.
- Verification: Generated code still has to be verified. Depending on the context, this is done by testing the output, by reviewing the generator and its templates, or by qualifying the tool. Safety standards set specific requirements for tools whose output is not verified separately.
- Documentation: Record the generator version, templates and input data for each build so a result can be reproduced. Documentation teams can use the same input data to generate tag lists, I/O references or parameter tables for manuals.
Automatic code generation vs. AI coding assistants
Classic generators are deterministic: the same input gives the same output. An AI coding assistant produces suggestions that can differ between runs and must be reviewed like hand-written code.