With AI and simulation providing more comprehensive insight on performance, designers can focus on how to optimize designs.

Hexagon has sponsored this post.

For decades, automotive designers have relied heavily on simulation to assess how various design parameters impact vehicle performance. As technology has advanced, simulations have become more sophisticated and realistic, and capable of more closely depicting actual conditions.

With the advent of artificial intelligence (AI), simulation is taking on a new level of sophistication and realism. AI-enabled simulations are allowing designers to conduct simulations more efficiently and precisely than ever before, potentially revolutionizing automotive design.

What AI offers

Simulations in the automotive industry involve a wide range of data, both physical and digital. One of the major challenges has been managing all that data — sorting through information to distinguish between what is significant and insignificant — an often tedious process. AI can help combine physical and digital data from various sources, identifying patterns and allowing designers to more efficiently extract intelligence from that data.

By helping identify correlations between various datasets, AI enables designers to gain new insight into improving vehicle design. For example, AI-enabled modeling can predict manufacturing costs for different designs, enabling manufacturers to select the most efficient design.

AI can also improve the efficiency of executing simulations. Traditionally simulation has required multiple experts in specific areas to develop models, analyze results and make adjustments for subsequent simulations. With AI, designers can develop algorithms with machine learning (ML) to understand design objectives and answer questions in real time without having to create additional models, dramatically reducing simulation time.

Along with the time savings, AI-enabled simulation allows a wider range of people to run multiple simulations using a trained model. It democratizes the information, so that anybody — within reason — can start to understand the effects of design changes on performance, explains Hexagon’s Manufacturing Intelligence division, which develops simulation, prototyping and other technical solutions for the automotive and other industries.

With AI providing more comprehensive insight on performance, designers can focus on how to optimize designs to meet certain objectives. Hexagon’s tools can target a particular set of outcomes within constraints of the design space and help find the appropriate inputs to get to a target outcome.

How it works

To develop AI-enabled simulations, teams must first identify, gather and prepare available performance data pertinent to the simulation. This includes categorizing and labeling model data, defining key variables and identifying desired outcomes. 

With a thoroughly prepared dataset, the next step is to train the AI model, using known physics and engineering data to guide the ML, model creation and model refinement. In some cases, this may also involve sharing model data with other sources to refine the model based on external data.

As the model produces results, it can be optimized using parametric, sensitivity and robustness analysis. The results can identify trends, sensitivity, gaps in data and other aspects that may warrant further refinement. Ultimately, the goal is to extract correlation and conclusions for optimizing designs and performing future simulations.

In some of these areas, automation and AI can streamline the processes. In fact, AI can actually be used to help automate the development of the model. For example, AI can help to package performance information, such as simulation data, physical test data, observed data, into more consumable pieces, irrespective of source. AI can also enrich data, filling gaps in available data.

Hexagon provides several solutions for performing AI-enabled simulations that combine physical and digital data. ODYSSEE (Optimal Decision Support System for Engineering and Expertise) employs predictive techniques to help engineers avoid the challenges of generating large amounts of data. With built-in tools to connect to engineering data sources, it aids in data orchestration, ML, analytics, digital twin model creation, customization and other key functions. ODYSSEE and other Hexagon tools are application-agnostic, meaning they are not restricted to data generated by Hexagon applications.

SimManager, a web-based simulation process and data management (SPDM) system, manages computer-aided engineering (CAE) and other needs specific to simulation. MaterialCenter serves as a scalable system that integrates with Microsoft Excel and other third-party CAE applications, enhancing the accuracy of simulations with appropriate material data to support informed decision-making. Hexagon also provides a variety of other tools and services to support AI-enabled simulation. 

Use cases

A growing number of companies are leveraging AI-enabled simulation to produce tangible benefits. Faurecia, an international automotive technology company, has used AI-enabled simulation to drastically reduce the time needed to analyze vehicle seating in crash conditions. The AI approach, which used ODYSSEE and SimManager, enabled the team to work with both physical and digitally simulated test data and reduce by 93% the time required to produce results when compared with previous methods. 

Other Hexagon clients have used AI-enabled simulation to address a variety of needs in automotive, aerospace and other industries.

Future possibilities

Looking ahead, AI is poised to dramatically change simulation in automotive design and other fields, with advancements occurring on an ongoing basis. In this rapidly evolving landscape, the integration of AI and ML in the simulation process isn’t just a trend — it’s quite transformative in itself.

As AI-enabled simulation continues to advance, automotive designers will undoubtedly find new ways to perform complex simulations and apply the results. The dramatic gains in efficiency and reduced time required to perform simulations will allow designers to focus on optimizing designs and other tasks requiring critical thinking.

Learn more at Hexagon Manufacturing Intelligence.