The increasing functionality of automation software in complex mechatronic systems such as construction machinery and the associated issues with code maintainability are a major challenge for companies to remain competitive in the global market. A major difficulty is that the software development in con-struction machinery often involves employees from different disciplines who have technological expertise about the process but little software background. Low-code platforms allow software to be developed via intuitive graphical in-terfaces even without extensive programming knowledge. Although such platforms have already found their way into the world of construction ma-chinery, the resulting programs are often difficult to understand due to the so-called scaling-up problem that occurs in case highly complex technical pro-cesses are implemented using graphical programming languages. In high-level language software, there are various assistance systems to minimize the complexity of the code already during programming. However, such ap-proaches are hardly available for automation software in mechatronics. This paper thus presents an assistantt that supports the programming of automation software on low-code platforms to reduce the complexity of the resulting code. Innovative approaches from static code analysis and machine learning are combined to enable predictions about software blocks to be used and op-timal assistance for the user. For the example of the low-code platform eDe-sign, a graphical programming platform developed by HAWE Hydraulik SE, it is shown how users of the platform can be assisted in creating maintainable, reusable automation software in the construction machinery sector.
«