Artificial Intelligence & Machine Learning in CAE Simulation and Manufacturing
Solving your engineering problems in real-time
Take two months of work and reduce it to one week. This is the power of AI/ML applied to CAE simulation.
Machine learning (ML) methods based on deep artificial neural networks (deep learning) have achieved tremendous success in many applications in many industries. Often branded and known as Artificial Intelligence (AI), these methods are able to provide accurate, data-driven process automation.
In the context of Computer Aided Engineering (CAE), AI has the potential to speed up the development of tools that allow non-experts to use sophisticated simulation capabilities (so-called 'CAE democratization') to increase employees' productivity, to optimize the computational resources required for the simulations, and to improve the product design process through new insights.
Webinar: Machine Learning for Real-time Parametric CAE Simulations, Optimization and Robustness with ODYSSEE.
View Now
ODYSSEE CAE
Structures
In Automotive, it's crash. In Aerospace, it's Bird Strike. In Consumer Electronics, it's the drop test. The list goes on and on … All these examples require vast amounts of small changes that ultimately result in different answers, and a different structural simulation has to happen for each one…this is where predictive modeling can come in. Imagine only analyzing 1/100th of the necessary use cases and leveraging ODYSSEE to get the rest of the answers you need immediately.
In Automotive, it's crash. In Aerospace, it's Bird Strike. In Consumer Electronics, it's the drop test. The list goes on and on … All these examples require vast amounts of small changes that ultimately result in different answers, and a different structural simulation has to happen for each one…this is where predictive modeling can come in. Imagine only analyzing 1/100th of the necessary use cases and leveraging ODYSSEE to get the rest of the answers you need immediately.
Computational Fluid Dynamics (CFD)
Aerodynamics, subsonic flow, hypersonic flow, powertrain performance, cabin environment, thermal dissipation, and cooling systems are a few examples of CFD applications that require analyzing complex thermal and fluid phenomena. All of these simulations require vast amounts of time and effort for every design iteration. Data based (predictive) modeling based on ODYSSEE technology can drastically reduce the amount of CFD simulation needed to fully optimize your designs.
Aerodynamics, subsonic flow, hypersonic flow, powertrain performance, cabin environment, thermal dissipation, and cooling systems are a few examples of CFD applications that require analyzing complex thermal and fluid phenomena. All of these simulations require vast amounts of time and effort for every design iteration. Data based (predictive) modeling based on ODYSSEE technology can drastically reduce the amount of CFD simulation needed to fully optimize your designs.
Acoustics
The shape of a rear-view mirror, subtly changing the sound the wind makes at different speeds down the highway. The varying designs for dampening sound in the automotive cabin. The sound an aircraft engine makes at different areas of the fuselage, at different RPM. The buzz of a drone overhead a busy walkway… These all require many, many different analyses at each incremental design change. Predictive modeling has the ability to vastly decrease the amount of acoustic simulation needed to arrive at your answers.
The shape of a rear-view mirror, subtly changing the sound the wind makes at different speeds down the highway. The varying designs for dampening sound in the automotive cabin. The sound an aircraft engine makes at different areas of the fuselage, at different RPM. The buzz of a drone overhead a busy walkway… These all require many, many different analyses at each incremental design change. Predictive modeling has the ability to vastly decrease the amount of acoustic simulation needed to arrive at your answers.
System Dynamics
When studying the dynamics of moving parts and how loads and forces are distributed throughout mechanical systems, product manufacturers often struggle to understand true system performance until late in the design process after vast amounts of simulation. Predictive modeling allows engineers to evaluate many more scenarios allowing for optimized designs that are first-time-right.
When studying the dynamics of moving parts and how loads and forces are distributed throughout mechanical systems, product manufacturers often struggle to understand true system performance until late in the design process after vast amounts of simulation. Predictive modeling allows engineers to evaluate many more scenarios allowing for optimized designs that are first-time-right.
Manufacturing
Imagine being able to establish cost indicators for tooling instantaneously instead of the traditional time-consuming approaches, by applying AI/ML for improved process monitoring, fault detection and control. Typical manufacturing process simulations, like welding or forming are laborious and require specialized knowledge. This doesn't allow most design and process engineers to explore all the options for a fully optimized process and thus they end with less-than-ideal designs. Predictive modeling can assist with real-time optimization of the manufacturing process allowing engineers to design first-time-right and first-time-fully-optimized.
Imagine being able to establish cost indicators for tooling instantaneously instead of the traditional time-consuming approaches, by applying AI/ML for improved process monitoring, fault detection and control. Typical manufacturing process simulations, like welding or forming are laborious and require specialized knowledge. This doesn't allow most design and process engineers to explore all the options for a fully optimized process and thus they end with less-than-ideal designs. Predictive modeling can assist with real-time optimization of the manufacturing process allowing engineers to design first-time-right and first-time-fully-optimized.
Materials
Engineered materials offer product manufacturers several advantages in terms of weight and performance; however, they also come with several challenges during product design when compared to traditional materials. Material intelligence applies artificial intelligence to materials. Both physical testing and virtual testing leads to an invaluable dataset. With artificial intelligence, both can be combined to create more value. ODYSSEE's predictive modeling can be used to accelerate the selection of the appropriate material for the design helping to reduce both the amount of time and resources required.
Engineered materials offer product manufacturers several advantages in terms of weight and performance; however, they also come with several challenges during product design when compared to traditional materials. Material intelligence applies artificial intelligence to materials. Both physical testing and virtual testing leads to an invaluable dataset. With artificial intelligence, both can be combined to create more value. ODYSSEE's predictive modeling can be used to accelerate the selection of the appropriate material for the design helping to reduce both the amount of time and resources required.
Autonomous
How will you trust a fully autonomous vehicle to drive your kids to school or your elderly mother to her next doctor appointment? The answer lies in analyzing the vast amounts of data coming from cameras, radar, lidar, etc. resulting in trustworthy real-time decisions made by that vehicle. One of the major obstacles in realizing this vision is filtering the vast amounts of data available. Separating and/or merging the real from the virtual data (signals) captured is invaluable and ODYSSEE can help achieve this much faster than traditional filtering methods.
How will you trust a fully autonomous vehicle to drive your kids to school or your elderly mother to her next doctor appointment? The answer lies in analyzing the vast amounts of data coming from cameras, radar, lidar, etc. resulting in trustworthy real-time decisions made by that vehicle. One of the major obstacles in realizing this vision is filtering the vast amounts of data available. Separating and/or merging the real from the virtual data (signals) captured is invaluable and ODYSSEE can help achieve this much faster than traditional filtering methods.
ODYSSEE A-Eye
Enabling real-time prediction using image-based machine learning for any industry.
Enabling real-time prediction using image-based machine learning for any industry.
- Images
- Sensors
- Scalars
- Labels
- Curves
- CAD
Democratization of ML based ROM using ODYSSEE A-Eye
ODYSSEE-A-Eye enables the creation of dedicated customizations for production/non-engineers using image-based prediction or based on CAD geometry (STEP).
ODYSSEE-A-Eye enables the creation of dedicated customizations for production/non-engineers using image-based prediction or based on CAD geometry (STEP).
We have been able to reduce the duration of this repetitive analysis and optimization of the door padding which is used for side impact energy absorption from five days to half a day thanks to ODYSSEE AI/ML/ROM solutions”
Dr Julien Tersac , Faurecia , France
At AUTOLIV we are concerned with safety of real humans (and not only dummies). This requires a yet challenging computing effort for evaluation of our safety solutions. ODYSSEE software is a real breakthrough and provides a very promising perspectives in order to reduce drastically computing time and optimize our designs”
Bengt Pipkorn (Director Simulation and Active Structures, Autoliv Research)
The approach in this study is unique, as it can capture the pertinent parameters influencing the casing buckling and the evaluation of the magnitude of each.”
James Njuguna, Professor of Composite Materials & Academic Strategic Lead, Research
ODYSSEE's artificial intelligence solutions open up new perspectives in Ariane 5's in-flight data analysis”
B. Troclet, Structural analysis senior expert, Ariane Group
Whereas finite element solution would have taken more than 1000 hours to achieve the solution for the full design space, [ODYSSEE CAE] was able to accurately predict the full response within seconds.”
Laike Misikir, Product Development Engineer, Ford Motor Company
[ODYSSEE CAE's ROM approach for decision making in 3D printing insoles] is now actually used by podiatrists that are already clients to the company that was partly funding the study.”
Michel Behr, Researcher, Gustave Eiffel University
