Design Engineer’s Weekly

Broaden your perspective. Expand your knowledge. Love your work.

Subscribe to Design Engineer’s Weekly 

Cover Story

A ChatGPT for Every Engineer

ChatGPT developer OpenAI hosted its first developer event this week to much fanfare, revealing several new products that include an improved version of GPT-4 and a new API for developers to build their own assistants.

But the standout announcement was GPTs: custom versions of ChatGPT that can be tailored for engineering or countless other applications without the need for additional coding. 

Pre-fabricated examples include tools for creative writing, math tutoring and trip planning, but the real goal of GPTs is to encourage users to create their own. There’s even a GPT Store in the works that will pay creators based on how many people use their GPTs. All of this has us wondering how engineers in various disciplines and industries might benefit from their own GPTs.

Here’s a simple use case: materials science. Earlier this year, engineers at University of Wisconsin-Madison used ChatGPT-4 to reduce the workload of extracting data from scientific papers by 99%. A materials science GPT could potentially go beyond mere information gathering to assist with material selection, analysis and even testing through integrations with materials testing software.

Then there’s the possibility of creating industry-specific GPTs without needing any knowledge of the API. Imagine an A.I.-powered aerospace assistant trained on aerodynamics and propulsion data that could assist with flight simulations, or a structural engineering assistant that specializes in structural analysis and earthquake engineering. Of course, these sorts of applications would require GPTs to be much better at engineering data calculations than they are today.

What about CAD integration? We’ve seen some early attempts at this in other enterprise software, such as in PLM with Siemens Teamcenter and in CAM with SprutCAM X. A GPT integrated with CAD software would mean a natural language interface; it could offer design assistance, verification and documentation; and it would make it easier to learn design principles and best practices with automated coaching. If it actually works, the gap between design intent and CAD software would be smaller than ever.

In the News

Starship Readies for Another Launch

SpaceX announced it’s targeting mid-November for the second flight test of its Super Heavy rocket and Starship upper stage. A notable change is the addition of a hot staging ring between the booster and the upper stage, which is riskier but will also increase payload-to-orbit capacity. In case you’ve forgotten, the first flight test ended spectacularly yet was deemed a success.

Superconductor Skepticism Soars

It looks like the final nail in the coffin for yet another too-good-to-be-true room-temperature superconductor claim from this past summer. The scientific journal Nature officially retracted the paper in which the claim was made, as requested by 8 of its 11 co-authors. This, in addition to the debunking of LK-99, will no doubt cast a long shadow over any future claims.

Satellites are Going Green

A collaboration between Kyoto University, JAXA (Japan's space agency) and NASA is aiming to put a prototype satellite made from magnolia wood into orbit early next year. Using wood in place of aluminum should yield the same strength-to-weight ratio while being less expensive and more environmentally friendly. This isn’t the first wood sent to orbit: an earlier experiment demonstrated that wood panels can survive unaffected in low Earth orbit for at least a year.

Tool Tip

CADGPT: An A.I. Assistant for Engineers

For a sneak peek at an A.I.-powered CAD assistant, check out CADGPT. It’s a bit like an intelligent search engine for design knowledge. It can tell you about the duties of an architect in Canada, generate a LISP program to delete selected layers and even prepare contracts for CAD services. 

Recommendations

Netflix. Hulu. NASA?

NASA is officially in the video streaming business after this week’s launch of NASA+, which “puts the universe at your fingertips'' via live coverage and original series. It’s free to watch and, better yet, free of ads. Not sure where to start? Try this great video on how the James Webb Space Telescope was designed.

Chopsticks Robot Moves Mountains (of Boxes)

Dextrous Robotics is working on a novel design for handling warehouse inventory and unloading transport trucks. The latest version of the chopsticks robot can lift 40kg boxes at up to 3gs of acceleration, which translates into as many as 2,000 boxes per hour.

The State of Diversity in Engineering

The results of a recent diversity, equity and inclusion (DEI) survey paint a complex picture of the design engineering workplace. While more than half of respondents admitted to witnessing or experiencing discrimination or bias, an even larger majority said they felt comfortable speaking up about DEI issues at work.

From engineering.com

The Heart with a Digital Twin

Two-time Olympian and winner of the women’s Boston Marathon in 2018, Desiree “Des” Linden has a lot of heart. Dassault Systèmes and TCS will find out just how much with the Living Heart Project, which uses simulation to create better models of the heart. An MRI scan of Linden’s heart will serve as the basis for a digital twin that will be used to predict her future performance.

History vs Direct Modeling in CAD

Is it always better to edit a sketch or a dialog box rather than changing part geometry directly? Also known as “ordered modeling”, history-based modeling has been the default in CAD for years, but in some cases, direct-edit modeling could be a better option.

SOLIDWORKS Sketching for Beginners

It never hurts to have a refresher on the basics, and it doesn’t get much more basic in SOLIDWORKS than sketching. This article goes through the essentials of this versatile and fundamental part of modeling, giving you everything you need to know to start sketching in SOLIDWORKS.

Poll Results

The Battle for your Job

Last issue, we asked readers to select the statement that most closely reflects their feelings about A.I.’s potential impact on their career. The top choices, each with 44% of the votes, were “A.I. could do some of my work, but I'd still have a job” and “I have no clue how A.I. will impact my job, and that scares me.”