Hexagon democratises engineering simulation further with first Image-recognition AI for Computer-Aided Engineering

Hexagon’s Manufacturing Intelligence division has today announced the release of ODYSSEE A-Eye, a new artificial intelligence (AI) tool that enhances Computer-Aided Engineering (CAE) for a multitude of industry solutions without the need for complex CAE preparation and simulation expertise.

ODYSSEE A-Eye applies advanced pattern recognition to images, pictures, video, and simple datasets so that they can be analyzed for machine learning (ML) applications in engineering simulation. The tool’s unique ability to match existing CAE simulation data to imagery and provide like-for-like predictions based on a posed problem makes it an invaluable tool for use in any engineering-related problem – design optimization, cost impact, or even how long a new product will take to produce without the laborious and highly-skilled process of preparing a model and defining simulation boundary conditions.

Applying ML processes to engineering problems typically requires expert knowledge and large amounts of training data to produce valid and reliable results, which leaves it out of reach for many smaller enterprises and non-specialists. By removing the need for complex data models and allowing the user to solve their problem by inputting widely-available CAD files, imagery, or scalar data and relating it to training data from Hexagon’s simulation solutions, the ODYSSEE A-Eye platform makes powerful Digital Twin capabilities available to designers, production engineers, operators, and other non-specialists. They can then make informed engineering decisions and explore problems interactively with near-real-time results.

Example applications include:

• Exploring how car wheel designs behave when impacting obstacles such as a curb or debris. Engineers can build a database of different configurations using nonlinear finite element simulations such as the design or number of spokes to understand the effect of various designs. Vehicle design teams can then use this to quickly understand the behavior of a wheel without any engineering or CAE knowledge based on only a 2D image.
• Predicting lift and drag coefficients for new aircraft wing profiles based on a 3D image of a new wing design, by building a database of just 16 wing profile simulations from the widely-used National Advisory Committee for Aeronautics (NACA) definitions. Typically, this process would take several days, and demand the time and attention of a CAE analyst and multiple simulation tools.
• A machinist or machine salesperson using an ODYSSEE A-Eye application to predict how long a part will take to produce with a given CNC machine tool and metal, using just the database and 3D Step file, capturing valuable process knowledge for others to better plan production and bid for projects. By applying manufacturing process simulation, the same process can be applied to predict dimensional tolerances or the strength of joinery.

Engineers without machine learning knowledge can use ODYSSEE A-Eye to develop their robust AI applications based on any particular problem that needs to be overcome, from optimizing the tire-tread design to fault-analysis of computer chips, and then make them available to others who need that knowledge. The new platform integrates with all of Hexagon’s CAE solutions, working seamlessly with customers’ existing processes and bringing AI to industries that may not have seen this as a feasible solution to their current design needs. Its accessibility means it can be used by companies who either do not carry CAE specialists or want the expertise they do have to solve other problems or perform final design validation. With ODYSSEE A-Eye, a single engineering expert is able to specify an application that would help progress a design, and then feed that to the product design team and operators to execute.

Roger Assaker, President of Hexagon’s Design and Engineering Software Business Unit, said: “AI is an increasingly valuable tool within design and engineering, helping push virtual engineering to the next level. It has the potential to shorten the time taken to complete labor-intensive design tasks that may have previously taken days or weeks down to minutes or hours without losing simulation fidelity. Furthermore, the user-friendly design of ODYSSEE A-Eye makes it simple to integrate into modern engineering practices, democratizing a highly advanced process for use by non-experts, and producing the results in a very accessible format.”

This content was originally published on the Hexagon MI website.