Design Engineer’s Weekly

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

Subscribe to Design Engineer’s Weekly 

Cover Story

What Has A.I. Done For Us Lately?

Following the anniversary of ChatGPT’s explosive release at the end of 2022, it seems fair to characterize 2023 as a bonanza year for artificial intelligence. As major media outlets release retrospectives on the private AI arms race (subscription required) and tech giants speculate about future risks, we want to ask a much more basic question: 

What has AI actually done for engineers?

Well, there’s been no shortage of announcements for AI upgrades to the engineer’s toolkit, Ansys and Autodesk most recently. There are also indirect benefits, such as using AI for materials discovery (subscription required) or to conduct automated inspections. And who doesn’t appreciate having ChatGPT or some other large language model (LLM) in their back pocket?

More broadly, we’re seeing AI impact everything from weather forecasting to scientific research to interspecies communication. Given its ubiquity, it’s no wonder that technology leaders voted AI as the most important technology of 2024 by a wide margin. That would also explain the new AI Alliance aimed at promoting open-source development. 

And yet, according to the U.S. Census Bureau, only 3.8% of businesses are using AI to produce goods and services. The adoption rate is considerably higher in the information sector at 13.8%, but it’s still far lower than one might expect in light of all the recent hype. One thing we do know is that generative AI struggles with engineering innovation, which is a crucial part of producing goods.

Still, it’s difficult to ignore the many exciting possibilities for AI. Perhaps the key to staying grounded is focusing on hardware over software. For example, researchers at MIT and Stanford have used AI to improve robot training and a roboticist at the University of Vermont has argued that embodying AI in robots is the key to unlocking its full potential. Terminator clichés aside, we might not see AI’s true capabilities until it expands beyond the chat window and into the real world.

In the News

Design Decisions May Dig Cybertruck’s Grave

It’s official: nearly four years after its first unveiling, Tesla’s controversial Cybertruck is finally shipping to customers. Of course, how many and how quickly are the billion-dollar questions. Experts have argued that the vehicle’s unique design could be its undoing because of higher-than-expected manufacturing costs, a lack of appeal to American consumers, or both.

Roomier EV Interiors Promised by Drivetrain Redesign

South Korean automakers Hyundai and Kia presented a new design for electric vehicle drivetrains. The Universal Wheel Drive system–Uni Wheel for short–puts main drive system components inside the wheel hubs. Specifically, the system replaces the constant velocity (CV) joints with a planetary gear configuration that puts the motor directly behind the hub, thereby increasing the space available inside the vehicle.

IBM Catches Up on Quantum Computing

Although the start-up Atom Computing beat IBM to the punch by more than a month with the world’s first 1,000-qubit quantum computer, Big Blue is catching up in a big way with the release of a new 1,000-qubit quantum chip called Condor. More significantly, the company unveiled another chip called Heron, which has only 133 qubits but makes up for it with an error rate five times lower than the previous generation’s. At this rate, the company expects we’ll see quantum-computing-powered simulations before 2030.

Recommendations

Career Advice from a Formula 1 Engineer

John Judd has been designing high-performance engines for decades. At one point, his company was on the hook to supply three Formula 1 racing teams, on top of design work for Formula 3000 and IndyCar. Looking back on his prestigious career, Judd offers advice for younger engineers, including this: “Don’t take yourself too seriously. For those lucky enough to be clever, it isn’t necessary. If you’re not, then you’ll look an even bigger fool.”

Dr. Seuss Meets Engineering

Ever heard of oobleck? Theodor Geisel, aka Dr. Seuss, coined the term in a book about a magical material that threatens to overwhelm a kingdom. You might be more familiar with the real world formulation of oobleck: cornstarch and water. A team of molecular engineers recently investigated the non-Newtonian behavior of oobleck, research that could lead to the development of customized non-Newtonian fluids with controlled viscosities for infrastructure and biomedical applications.

From engineering.com

Faster Simulation with Siemens’ A.I.

Siemens is getting into the AI-powered simulation game with HEEDS AI Simulation Predictor and Simcenter Reduced Order Modeling. One notable feature leverages digital twins to address simulation AI drift, which can occur when models are asked to extrapolate data based on inputs that are outside the scope of the training data.

Configuring Workflows with Autodesk Fusion 360

Configurations in Autodesk Fusion 360 have the potential to streamline design optimization, reduce risk and improve efficiency. Autodesk senior manager Trent Still likens configurations to highways without unnecessary detours. With Fusion’s AI integration, there are even more possibilities for enhanced product development and customization.