Abstract
This paper describes in detail the research methodology, including the use of a local SQLite database to store information about drawings, materials, processing methods, and formulas. The need for dynamic generation of input fields, real-time calculations, and display of drawing thumbnails is emphasized. Particular attention is paid to the fault tolerance of the system, including exception handling (division by zero, missing variables, incorrect data entry) and ensuring stable operation without abnormal termination. The application architecture is organized according to the modular principle (interface, database, computational engine), which simplifies support and extension. The use of the eval() function in a strictly limited context for safe execution of formulas is noted. The application is developed in Python using PyQt5 and can operate autonomously, without access to the Internet or external servers. The presented work demonstrates a solution to the current problem of automation in mechanical engineering, offering a flexible, reliable and convenient tool for engineers and technologists