Automated code development based on genetic programming in graphical programming language: A pilot study

Area of Science:

• Computer Science
• Software Engineering
• Automation

Background:

• The automation revolution has increased computer technology's industrial impact.
• Software development and testing are time-consuming, with a shortage of experts.
• Automation can enhance software engineers' workflows and reduce time-to-market.

Purpose of the Study:

• To conduct a proof-of-concept for a requirement-based automated code-developing system.
• To implement automated code generation in the graphical programming environment LabVIEW.
• To address the gap in research concerning graphical programming environments for code generation.

Main Methods:

• Development of a requirement-based automated code-developing framework for LabVIEW.
• Evaluation of the framework on four benchmark problems: string, numeric, boolean, and mixed-type models.
• Testing fundamental programming scenarios to assess system capabilities.

Main Results:

• The algorithm successfully generated functional and errorless solutions meeting all user requirements.
• Generated code, while more complex, maintained execution speed and accuracy.
• Demonstrated feasibility and viability for automated code development in LabVIEW.

Conclusions:

• The pilot study proves the concept of requirement-based automated code generation in LabVIEW.
• The system shows promise for solving linear and binary programming tasks.
• This research opens avenues for empowering application developers and non-programmers.