November 30, 2023


Sapiens Digital

HxGN LIVE Design & Engineering 2021

For years, we’ve been hearing about how the next big industrial revolution is either well underway or about to get started in earnest. Both of those statements have some truth in them, but the former is starting to look more and more like it’s the horse to back. 

Industry 4.0 is, in essence, the bridging of physical and digital systems, achieved through the industrial Internet of Things, that is changing manufacturing as we know it. These new, artificial intelligence-enabled modes of testing and production are beginning to help engineers monitor physical assets in real time, set up autonomous decision-making systems throughout production lines, and enhance value creation processes.  

The revolution is so promising that the U.S. Senate voted to approve a bill that has been called the most expansive industrial policy legislation in the country’s history back in June. And, while it’s fair to say that most applications of Industry 4.0 initiatives still fall within the scope of systems from the last industrial revolution, genuinely exciting technologies and applications are beginning to emerge, most of which have a basis in AI. 

Hexagon’s Manufacturing Intelligence division has been a part of this change for years. The company’s focus on incorporating intelligence into design, engineering, production, and quality inspection solutions for their clients’ factories has earned them a rightful place as a prominent figure in any Industry 4.0 conversation. 

Fitting, then, that they’re hosting the HxGN LIVE Design & Engineering 2021 virtual conference over three days starting October 12. The event will see 220 talks on topics ranging from acoustics to ICME (integrated computational materials engineering), multi-physics digital twins, and AI and machine learning and how these fields are adapting to and leading the new industrial era. 

How AI solves a computational scaling problem

It’s worth reminding ourselves how we got here, as the talks we’re about to mention all play out against the same contextual industry backdrop. FE (finite element) analysis enables engineers to build high-fidelity, virtual representations of a physical asset and see how it will perform under a range of circumstances. 

These virtual simulations are often faithful to physics but also significantly time-consuming and computationally costly. Using these simulations alone is unlikely to help engineers realize the dream of digital twins, for example, one of the goals that characterize the current industrial movement. 

Digital twins are exact virtual representations of products, people, whole processes or even supply chains, offering both a live window into what is happening to them and how they might respond to varying situations over the course of their lifespan. By combining real-time data from sensors with the predictive power of simulation, digital twins help engineers make better-informed decisions with greater speed, allow for more comprehensive approaches in research and design, and simplify complex optimization problems and product innovation. 

Two men use a computer-aided design program to analyze a machine part.
Source: Kumpan Electric/Unsplash

Realizing these requires help from AI, however, as FE simulations have so grown in complexity in recent years that they present real computational and financial obstacles for the engineers and designers who need to use them. AI changes that. Hexagon, for example, has already seen success with CAE-aware machine learning, using it to create real-time simulations to test automotive hardware, something that’s just impractical with physics-based simulation alone. 

AI gets around that impracticality by providing the information needed to make a ROM (reduced-order model), a mathematical approach to identifying a traditional model’s defining traits and preserving them in a simplified, computationally efficient way. Machine-learning algorithms achieve this by “studying” the dynamic between input and output values of past FE simulations.

This is a highly valuable tool in a world where FE simulations of complex systems like those in CFD (computational fluid dynamics) can take hours or even days to run. AI can bring that timing down to seconds. It also helps lay the groundwork for digital twins, whose real-life accuracy will hinge on how well researchers can train the algorithms that come before them. 

Using AI in manufacturing spaces can democratize highly complex engineering tools to the practical advantage of non-experts. It also allows for a wider range of collaborative possibilities without risking intellectual property, as ROMs by their nature often obscure proprietary information. 

HxGN LIVE Design & Engineering 2021: Talks to watch

There are hundreds of talks taking place over the course of the conference, and all on fascinating subjects, so it’s difficult to single out any number of them for a highlight reel. But if you’re curious to see how AI is helping to shape manufacturing, design, and Industry 4.0 at large, we’ve put together a short list of a few presentations you’d be remiss to miss.

The automotive manufacturing industry is among the most receptive AI playgrounds out there. As the electric vehicle (and to a lesser extent, the self-driving) revolution starts picking up a global pace, the automotive world is reinventing itself with AI, even using it to better understand the ergonomics and safety considerations of seats under both normal driving conditions and crash scenarios.

Building and performing the simulations for both of those conditions can be quite time-consuming, especially given that they are often run hundreds of times over while considering varying postures of occupants to determine injury values. 

JSOL Corporation representative Masahiro Takeda will be giving a talk regarding a case study that attempts to predict such injury values by linking driving and crash simulations with ROMs generated in part with machine learning algorithms. Takeda’s study is a great example of how ROMs can alleviate computational load, as their ROM’s average prediction accuracy came in at above 90 percent and sliced what would have been a 19-hour computer crunch into a calculation of only a few seconds

Laurent Di Valentin, CAE Senior Technical Fellow for Stellantis, will deliver a talk on machine learning and CAE-complementarity, detailing how the company is building behavior models using artificial intelligence to augment and replace more traditional physics models for vehicle project development. These models increase efficiency all the way from the design stage to validation and are an illustration of how AI can expedite the whole system. 

Another talk centering on the synthesis of FE simulations and machine learning is Gustave Eiffel University researcher Dr. Michel Behr’s presentation on the real-time design of 3D-printed, orthopedic insoles. Behr has been involved in impact biomechanics and injury prediction for the last 15 years, and his talk will focus on the development and testing of new ways to predict how insoles will affect a patient’s gait. 

A woman in a factory looks at a computer-aided design of a car.
Source: ThisIsEngineering RAEng/Unsplash

The results of Behr’s work indicate how beneficial it may be to combine the best of model reduction methods and more traditional FE modeling to create better decision-making tools for designers and planners. 

One of the more topical presentations we’re looking forward to is Hideki Nakata and Zhuravlev Anton’s talk on using CFD and AI to better understand and prevent the airborne spread of infectious diseases like COVD-19 as well as better preparing cities to respond to natural disasters. 

Using St. Mary’s Cathedral in Tokyo as a test space, researchers at Ecokaku Kyoto’s Technology Division, in collaboration with Hexagon’s Leica Geosystems team, developed a droplet visualization system in which they were able to conceptualize the airflow (and thus the potential range of the virus’ spread) inside the building. 

Leica’s 3D laser scanners and Hexagon’s Cradle CFD software enabled them to construct a model of the building’s interior, visualizing how its architecture and air conditioning systems would distribute the airborne water particulates that are the most common vector for transporting the virus.

Normally, such modeling can take a considerable amount of time, but Ecokaku was able to utilize AI to convert the analyses of measured data into a ROM that allowed for real-time views into a number of “what-if” airflow scenarios. Doing so helped researchers develop effective countermeasures to reduce worshippers’ risk of exposure to the virus, allowing the surrounding community to gather for Christmas day celebrations. 

Intriguingly, the techniques used in modeling the cathedral space and developing safety countermeasures are also being applied to the analysis of extreme weather events. Organizations like ecoKaku can use the same AI tools to provide actionable plans to city officials and engineers to better prepare for and respond to natural disasters. As climate change promises an increase in extreme weather events in the immediate future, AI-aided analyses like these could very well end up saving lives. The talk promises to be noteworthy. 

Finally, Leonardo Aircraft representative Stefania Sorrentino will be giving a presentation on materials database enrichment using AI. Sorrentino will discuss how machine-learning techniques can be used to predict the behavior of a material — in this case, composite layups — without actually running material behavior simulations for all, let alone destructive testing of the material and the cost and time that requires. 

These predictive models could potentially be extended to other areas of design as well, enabling researchers to better decide what tests need to be run to improve the model’s accuracy and how to best combine these newer AI tools with traditional virtual testing. 

Building instead of predicting the future of AI

As the world of design and production moves further and further into Industry 4.0 practices, Hexagon is holding an aptly-timed conference, especially given how these technologies are almost guaranteed to soon alter how pretty much every manufacturing sector goes about its business.

Just as in every other field it’s being applied to, AI is playing a key role in that transformation. The fourth industrial revolution still bristles with possibility, and predicting just how technological tools like this will change the future is a big part of what we do here at Interesting Engineering. But the hundreds of speakers at HxGN LIVE Design & Engineering 2021 are the ones actually shaping that future. Don’t miss out on what they have to say about it.

Source Article