The arrival of user-friendly large language models (LLMs) has driven AI hype to deafening levels in the past few years, but many engineers have long been integrating AI and machine learning into their work. A new breed of engineering-focused AI tools is now combining the best of both worlds.
Here are some of the leading tools being deployed by engineers, including new entrants, long-established favourites and new functionality for stalwarts.
Civils.ai
Developed by a team of civil and structural engineers, Civils.ai extracts and organises construction project data from things such as site reports, contracts and CAD.
The AI then sorts the data and allows engineers to search for and summarise project information through the program’s own search engine, as well as feed back information to external databases.
Transcend
Transcend uses AI-powered generative design to help engineers quickly create designs for infrastructure projects such as water treatment facilities and power transmission and distribution assets.
The cloud-based program allows users to run multiple conceptual designs for multiple sites, and also generates a wide range of detailed, technical documents including single line diagrams, process schematics, BIM and 3D models, among others.
WolframAlpha
Symbolic computation engine WolframAlpha has long been a go-to AI engine for the STEM industry. Unlike ChatGPT, which is based on statistical approaches to training large language models (LLMs), Wolfram computes factual information with “symbolic AI”. In short, it tries to learn in the same way as humans do.
It can be particularly useful for checking complex mathematical equations and creating corresponding charts and graphs.
Last year, Wolfram became one of first plug-ins for ChatGPT, vastly improving the latter’s usefulness for non-trivial calculations and for producing verifiably correct answers. The plug-in allows users to use ChatGPT to restructure their queries to get the best out of Wolfram. This post from Wolfram’s founder explains it well.
Leo AI
Leo AI is a generative AI platform for mechanical engineers that helps turn text and sketches into 3D models. Composed of three interconnected modules – Leo Ideation, Leo CAD and Leo Drawing – each module feeds into the next to create concepts from scratch or improve existing designs.
Its reference library also helps to answer technical questions, such as material properties, thermal treatments and even help coding dilemmas. Leo already has the capability to describe a product and have it generate visual concepts with a technical sheet, with its next evolution set to describe the functionality or geometry of a component and receive an editable CAD part with a feature tree.
Synera
An automation platform specifically designed for engineers, Synera’s clients include NASA, which uses the program as part of its “text-to-spaceship” generative design process. A low-code software that uses a node-based interface, it helps to automate repetitive engineering tasks and save them as templates for re-use.
Synera also connects more than 20 CAD, meshing, analysis and productivity software tools, as well as allowing users to add in their own Python scripts.
TensorFlow
A free and open-source software library for machine learning and artificial intelligence, TensorFlow has a passionate community of developers, scientists and engineers who make it tick.
Often described as the most popular AI engine thanks to its powerful and mature library, TensorFlow is known for its strong visualisation capabilities and options for high-level model development. It can be used with frameworks such as Keras, an open-source Python library that provides a high-level interface for the TensorFlow library, allowing engineers to more quickly develop deep-learning models and neural networks.
Heuristica
Primarily designed for R&D engineers and tertiary researchers, Heuristica is a ChatGPT-powered program for engineers wading into new territory. It’s an AI-powered concept mapping tool with a user interface that looks like a mind map, where users can go deep into subjects by clicking on areas of interest — without prompt engineering. A user can take an unfamiliar concept and begin creating a visual map with features such as significance, implications and examples to broaden their understanding.
Coding assistants
For those in the market for a simple productivity boost, AI-powered coding assistants can automate relatively simple and repetitive tasks such as code generation and completion, as well as streamlining code and checking your homework. Some engineers find that ChatGPT does the trick, but others prefer paid programs such as TabNine, which learns from the user’s coding style and improves with use.
Summarising tools
They don’t say it loudly, but even academics are using PDF summarising AI tools to help them extract key information from papers, textbooks and technical manuals. Claude is considered the gold standard in this field, as it is less prone to “hallucinating” ideas or skipping over important information, and can digest larger reams of data.
If it’s a digest of an audio or video that you’re after, Summiz can help.