Designs led by simulation has changed the product development process, it allowed engineers to work on fewer design iterations by conducting virtual tests and thereby saving costs in producing fewer prototypes. A tool that has now become an essential element to the product development process. Over time, the advancement of computing capabilities allowed the possibility of conducting complicated analysis and the ability to solve design problems with better accuracy and precision.
There still lies certain shortcomings in the way simulation tools perform, where complex analysis lead to longer development cycles, thereby consuming more resources. Thus, the entrance of Machine Learning and AI enabled software to once again transform the world of physics-based simulation.
Existing CAE model simulations serve as the basis of information for the creation of datasets containing the input data and the respective output data. Multiple simulations with the variation of design parameters or predictors – which define the system under simulation, provide numerous results in the change in nature of the system under observation. These results train a Neural Network (NN) to create its own algorithm which can be used to predict results for test data to verify the deviation of results between the neural network and traditional CAE process. The larger the variety of simulation trials, the better the model learns to the information and is less error prone. The trained network is then used to predict results for new data at a significantly quicker speed compared to the conventional simulation process.
AI augmented CAE is heavily dependent on the data that is provided to train the neural network for its application, it reflects results based on the type of data (clean or tainted) and hence is unable to tell results from right or wrong, it solely relies on the dataset it is trained on. As for the accuracy of this technology, with increasing number of simulation iterations, more data is fed into training the model, and the algorithm it generates continues to improve and refine itself to reduce the error rates in the final predicted output. The network is hence improving its accuracy under supervised learning continuously. Good trained models are noticed to have very low error rates (1~10%) and are significantly quicker to the CAE simulations that they are based on. Even with the error rates, the industry is inclined to use this technology because of the amount of time and resources that are saved in doing so, and hence is seen as a viable and important part of the simulation industry.
Although AI has been actively used and implemented in various industries, it is seen to be in the nascent stages in the simulation industry. Implementing AI still requires a large database of performed simulations for it to be functional, and its accuracy and usage will only increase over time as more simulations are carried out to provide a larger dataset, which always takes time and resources to set up.
With the application of AI, simulation engineers must also be aware and have a basic understanding of how algorithms work for them to implement an AI driven model with their simulation studies. Data engineers and simulation experts are required to work together in order to achieve such understanding and clarity when it comes to offering AI augmented simulation as a service, a partnership which was not commonly seen earlier.
The benefits however are abundant, implementation of AI augmented simulation will allow a much quicker product development process and thereby reducing CAD modeling and CAE software costs. Prediction of results could also be done concurrently with CAD development to make the process of CAD to CAE more seamless. Certain design iterations can also be carried out that were thought to be difficult to analyze and test, thereby leading to new product innovations.
KPIT is working on integrating machine learning with product development in order to
accelerate the development process along with significant cost benefits. Initial applications in
steering systems, chassis, drive axles and overhead crash guards have demonstrated direct
savings in cost and time, along with other intangible benefits. Other applications also include
electronics thermal predictions, seat comfort predictions and automatic dimensioning of
AI – Based Development Process at KPIT
1. Develop Model Database
2. Identify Machine Learning Features
3. Train Model using Available CAE Results
4. Validate and Quickly Predict Future Designs
1. AI-augmented Simulation (fraunhofer.de)
2. Augmenting CAE in Automotive design with AI tools – Intuceo
3. Data Driven CAE | EinNel Technologies
4. Simultelligence | Artificial Intelligence | Siemens Global
5. Can AI Take Simulation to a New Level? – Digital Engineering 24/7 (digitalengineering247.com)
6. The Coming of Age of AI and Machine Learning in Design – Digital Engineering 24/7 (digitalengineering247.com)
8. AI Driven Solutions for Cost & Time Reduction – KPIT
9. AI, Augmented Intelligence | Punit Dhillon
10. Integrating Artificial Intelligence with Simulation Modeling – AnyLogic Simulation Software
Connect with us
KPIT Technologies is a global partner to the automotive and Mobility ecosystem for making software-defined vehicles a reality. It is a leading independent software development and integration partner helping mobility leapfrog towards a clean, smart, and safe future. With 11000+ automobelievers across the globe specializing in embedded software, AI, and digital solutions, KPIT accelerates its clients’ implementation of next-generation technologies for the future mobility roadmap. With engineering centers in Europe, the USA, Japan, China, Thailand, and India, KPIT works with leaders in automotive and Mobility and is present where the ecosystem is transforming.
Rajiv Gandhi Infotech Park,
Hinjawadi, Pune – 411057
Phone: +91 20 6770 6000
Frankfurter Ring 105b,80807
Phone: +49 89 3229 9660
Fax: +49 89 3229 9669 99
KPIT and KPIT logo are registered trademarks | © Copyright KPIT for 2018-2023
|cookielawinfo-checbox-analytics||11 months||This cookie is set by GDPR Cookie Consent plugin. The cookie is used to store the user consent for the cookies in the category "Analytics".|
|cookielawinfo-checbox-functional||11 months||The cookie is set by GDPR cookie consent to record the user consent for the cookies in the category "Functional".|
|cookielawinfo-checbox-others||11 months||This cookie is set by GDPR Cookie Consent plugin. The cookie is used to store the user consent for the cookies in the category "Other.|
|cookielawinfo-checkbox-necessary||11 months||This cookie is set by GDPR Cookie Consent plugin. The cookies is used to store the user consent for the cookies in the category "Necessary".|
|cookielawinfo-checkbox-performance||11 months||This cookie is set by GDPR Cookie Consent plugin. The cookie is used to store the user consent for the cookies in the category "Performance".|
Leave a Reply