AI assistant or rival to engineering professions?

The main articles are:

• Artificial intelligence (AI, Artificial intelligence, AI)
• CAD Computer-aided design systems

One of the promising areas is the introduction of AI systems into processes that are previously completely dependent on humans: creativity, creative professions or the replacement of engineering work with artificial intelligence. Automation of engineering activities through the use of such technologies has been discussed for several years, and a lot of efforts have been directed into this area. However, despite the achievements, the bulk of the project work is still done by the individual. On October 30, 2025, expert Klyuchnikov Bogdan told TAdviser whether it is possible to replace or significantly automate engineering work using AI.

What is the essence of engineering? The main task of the engineer is to create new or optimization of existing structures and solutions based on technical knowledge, the laws of mathematics, physics and other disciplines. He finds out how things work, and finds practical application to scientific laws and discoveries in order to create a product that performs specified functions and solves the customer's problem. The key here is the engineer's idea of ​ ​ "For what? What do we solve the problem? ": Regardless of the appearance of the final product, it must fulfill its main function and meet the requirements. A common misconception is that engineering work is primarily the preparation and execution of technical documentation in the form of drawings or 3D models. In fact, this is the final design stage, expressed in the design design of the results of the design process, which includes various modeling methods and calculation methods.

𝓲

Фото: Freepik

The engineer translates the customer's needs (assigned task) into a technical solution (a set of product functions), and then details it to the level of individual components. The purpose of the process is to develop a set of detailed requirements, satisfying which, you can implement a design that solves the problem (performs a function) at its level. Collectively, the development of solutions to single problems (functions) by each component should close the customer's task. Systems engineering - a discipline that formalizes the approach to creating complex systems, emphasizes the special importance of requirements engineering.

This is the process of identifying more detailed lower-level requirements based on current-level requirements that determine what functions the product as a whole and its individual components should perform. It is at the stage of decomposition of requirements "top-bottom" that "engineering magic" takes place: the transformation of the designation of need into a specific technical solution, which will be reflected in the engineering documentation.

To find solutions, the engineer uses a variety of tools - diagrams, diagrams, sketches and methods for calculating the impact of various physical effects. Modern tools for automating engineering tasks allow you to design mechanical, hydraulic, electrical systems and test their work, and applications for 1D modeling provide the ability to automatically calculate system models. Such analytical methods of modeling and calculations are aimed at determining the proper functioning of the developed component or the entire product. The result is a clearly formulated set of initial technical requirements for each component, which will reflect the capabilities and operating conditions of the product - an accurate technical task for its implementation in physical form.

The next step is to move from the requirements that reflect the functional features of each element for the component being created to modeling the physical components. Here, the engineer works with functional elements from which you can now assemble the product design. This is another important point: despite the fact that in the end the structure is presented as a set of geometric bodies, for an engineer it looks like a set of functional elements - bosses, supports, levers, brackets, links, earrings, stiffeners, eyes, bearing seats, saddles, cams, etc.

Consider the example described above on the example of the simplest mechanical system - cups. Based on the main need (task) of the customer - to hold the vessel in the hand and use liquid - it is necessary to develop a technical solution that ensures the fulfillment of these functions. Such a solution is a glass, the shape of which can be arbitrary - cylindrical, prismatic, conical, etc. - provided that it is conveniently placed in the palm. The form is not critical, since it does not affect the basic functionality. If a new need is added to the initial task, for example, safety when interacting with a hot liquid, then a new function is required - to hold the vessel in such a way as not to burn the hand. In this case, the technical solution may be the presence of a handle, the shape of which is not an important factor. Thus, when new requirements are added (for example, the convenience of transportation in a bag, the ability to drink liquid without wetting a mustache, etc.), a new function appears each time, the implementation of which also does not depend on the specific shape of the object. Analyzing the forms with which the engineer works, it can be noted that they are a rational way to achieve a technical solution: least weight, stability, etc. Aesthetic aspects can also be considered, but their significance is determined by a specific task.

When analyzing various mechanical systems, including the simplest mechanisms, you can find similar approaches to solving problems. For example, the piston must be moved along a specific rigid axis, so this is limited by the presence of a cylinder. Lengthwise displacement force transmission is increased through rocker with different length of arms. Thus, it can be confidently stated that the engineer in his work is mainly focused on the functional aspects of the solution and pays less attention to geometric modeling. This is what often causes failures in reverse engineering projects, when any component is restored from existing geometry, but it does not provide the characteristics that the original object possessed. This is due to the lack of necessary calculations and non-compliance with manufacturing techniques that are directly related to the context of the application of this component, although the geometric shape is identical.

As of October 2025, a direction is actively developing related to the development of AI technologies for generating geometry based on 3D scanning data (point clouds), photo images from different angles. It is worth noting that tasks of this kind have already been considered by CAD system developers, since for some industries the process of developing the product style implies the presence of a physical model that designers create. The applications developed for this could successfully determine the geometric primitives from which the final shape was obtained, which is necessary to recreate the exact geometric representation, based on the point cloud. These tools successfully solve the problem of transforming design solutions into an end product. However, the recognition of functions that a particular geometric set performs was beyond the scope of the tasks.

If you present the engineering process as a whole, you can note the following steps:

• Demand analysis
• schematic solution model (diagram, sketch);
• calculations (multiphysical models);
• development of detailed requirements for solution implementation;
• collection of engineering (functional) elements for integration into a physical solution model;
• validation.

At the stage of formation of the principal solution model, there is an extensive set of template options described in the engineering reference literature. You can also use generic designs to create a set of engineering (functional) features and associate them with the functions you perform. As for the validation stage, there are currently quite effective means for carrying out multifysical engineering calculations and spatial analysis. Thus, the main problem of automation of engineering work is to link the above steps into a single cause and effect chain and the ability to discard inappropriate options.

A natural question arises: what areas of application of AI technologies may be the most in demand and how to benefit in the near future? In terms of prospective development, the focus of AI training is to focus on establishing a function-to-engineering relationship rather than generating finite geometry. Each part of any product is a set of functions and properties that meet very specific needs. Another area may be related to the use of archives of industrial enterprises, which contain a huge number of different engineering solutions, to create new products or develop existing ones. Many of them have extensive funds of drawings reflecting decades of engineering experience. In the process of implementing product data management systems or document management systems, many enterprises digitized archives to speed up the search and receipt of scanned copies of paper drawings, but the qualitative effect on borrowing technical solutions when developing products using 3D modeling was small.

It is characteristic that the drawings have fairly clear standardized design procedures, which makes them a valuable data source for AI training and content recognition, from which it would be possible to restore the 3D model of components in a fully automated mode. This will not replace the work of an engineer, but such an opportunity will give a significant impetus to companies to switch to full-scale design in 3D space and gain the benefits of this approach - from spatial analysis and simulation of behavior to validation using engineering calculation systems and the release of interactive materials for operation and training of personnel. Of course, the formation of an extensive array of data, supplemented by information about functions associated with engineering elements, should become a key basis for the training and development of artificial intelligence technologies in the engineering field. This will create the prerequisites for building intelligent systems that, although they will not completely replace the engineer, will be able to act as his reliable and indispensable assistant, significantly increasing the efficiency of design solutions.