ÂÌñÏ×ÆÞ

In the current competitive product development landscape, AI and machine learning (ML) prove a crucial role in enabling engineering teams to virtually validate design performance before physical prototypes are built — speeding up the process while improving accuracy. (Bangalore, India) seeks to maximize the impact design exploration with generative AI and accelerate CAE simulation through its platform, Xitadel Intelligent Process Automation (XIPA). Powered by ML, XIPA streamlines the CAE model-build process with high speed and reliability. 

XIPA is a proprietary ML system that combines Image Recognition, Object Detection and Classification to interpret CAD data and streamline finite element (FE) model creation. Using convoluted neural networks (CNN), it performs intelligent feature recognition to automatically detect low-level geometric entities essential for connectivity and fastening in CAD models, leveraging a geometric feature library built through more than 20 man-years of training, particularly for plastic components. These detected features are then classified into families, enabling tailored meshing and processing logic, after which Xipa applies customized meshing algorithms to rapidly generate high-quality FE meshes optimized for plastics and sheet metal, compliant with both industry standards and client-specific requirements. It further automates connection assembly for seamless integration during model building and generates multiple mesh outputs in parallel — such as crash; noise, vibration, harshness (NVH); and durability meshes — ensuring models are simulation-ready across attributes.

Xipa delivers significant efficiency gains by reducing model-build times — Xitadel reports up to ~50% faster turnaround than traditional methods, with customized ML models offering reductions of more than 80%. It ensures outputs are consistent with user-defined modeling guidelines and best practices, while also embedding organizational knowledge through ML training to capture and reuse institutional expertise for consistent modeling across teams.

In addition to speed and consistency, Xipa simplifies processes with “push button” automation via a user-friendly GUI that enables near-batch-speed meshing and assembly of complex subsystems, making virtual product development more accessible and efficient. Beyond pre-processing, it supports the entire CAE workflow — from load case setup and automated deck preparation to AI-driven postprocessing, predictive analytics and gen AI-based assistance — offering an end-to-end solution for simulation.

Summarily, Xipa significantly accelerates the time traditionally taken for meshing and model-build. It is therefore is an enabling technology that is on the critical path for the digital transformation of the model-build phase of product development process.