AI and Computer Aided Design 

The aim of this article is to gain an understanding of the myriad ways in which AI could become integrated with CAD. It is firstly important to appreciate that this is a nascent technology, and that at this point in time many of the opportunities are from “hacking” together AI tools, which were not designed for this purpose, but are effective nonetheless. In the future we can imagine they will become integrated in CAD specific software.  

CAD, or Computer-Aided Design, is a digital technology that enables the creation, modification, and optimisation of designs for various parts, structures, or systems. Using software, CAD allows engineers, architects, and designers to produce detailed and accurate two-dimensional (2D) or three-dimensional (3D) models of their concepts. 

Invisible AI:  AI's impact will likely be mostly behind the scenes, and many of its contributions are invisible to end users. AI can assist in generating code snippets or even entire programs based on high-level specifications provided by developers. This helps streamline the coding process and reduces the amount of manual coding required. In data analysis, AI algorithms can identify patterns and trends in large datasets that might be challenging for humans to discern. AI can automatically detect unusual patterns or outliers in data, helping identify potential errors. Algorithms can analyse CAD models for potential errors or inconsistencies, ensuring that designs meet specified standards. This helps in preventing issues such as geometric errors or manufacturability issues, providing a higher quality end product. AI can also facilitate predictive simulations, allowing designers to anticipate how a design will perform under various conditions. This can include structural analysis, thermal simulations, or fluid dynamics, contributing to better-informed design decisions without the end user directly engaging in the simulation process.

AI as a workhorse: AI may also be able to contribute significantly to CAD as a “workhorse” by automating repetitive tasks, streamlining complex CAM processes, and cleaning up 3D scan data. In Drafting and Annotation AI-powered tools may be able to automate the creation of 2D drawings by extracting information from 3D models, and vice versa. This includes generating annotations, dimensions, and other details, saving designers from manual and repetitive drafting work. When linking to CAM, AI algorithms may be able to optimise toolpaths for CNC machining, taking into account factors like material properties, cutting forces, and machine capabilities. This optimization enhances efficiency and reduces the likelihood of errors in the manufacturing process.  AI will perhaps be able to analyse the geometry of a part and automatically select the most efficient machining strategy; this includes determining the sequence of operations and tool changes for complex components. AI techniques, such as deep learning, can be used to reconstruct smooth and accurate surfaces from imperfect or incomplete 3D scan data. This is particularly useful in reverse engineering applications, but with copyright and IP issues to consider - can designers add DRM to a geometry for example?

AI as a tutor: 

AI tools can function as effective tutors in learning and using CAD by providing guidance, assistance, and suggestions throughout the learning process, e.g. Fabrio CAD Assist. AI can generate interactive, step-by-step tutorials for CAD software. This helps beginners understand the software interface, tools, and basic commands by providing guided instructions and visual demonstrations. AI can offer context-sensitive guidance, providing relevant information and tips based on the user's current actions within the CAD software. This helps users learn in a more personalised and adaptive manner. AI with NLP capabilities can enable users to interact with CAD systems using natural language. Users can ask questions, seek guidance, or request information in a conversational manner, making the learning experience more intuitive. AI can understand and respond to voice commands, allowing users to perform actions in CAD software through spoken instructions. This hands-free approach can be especially helpful for users who prefer auditory learning.

AI tools have the potential to break down discipline silos between designers and engineers by fostering collaboration, facilitating communication, and providing a unified platform for multidisciplinary work. For example, AI can integrate design and simulation tools, allowing designers and engineers to seamlessly transition between creating and testing. This integrated approach encourages a more iterative and collaborative design process.

At the time of writing, many AI tools are discreet, but designers are still gaining advantages by exploring ways in which they can be “hacked together.”  For example, if a student is designing in Tinkercad they have limited rendering capabilities. By exporting images and using tools like NewArc.ai they can quickly produce photo-realistic renders, sometimes in seconds. The dominant CAD platforms are already starting to roll out AI powered features and their power and functionality will inevitably only grow.