BMW Group boosts production efficiency with NVIDIA DGX and generative AI

Overview

BMW Group uses NVIDIA DGX systems to train deep-learning-based synthetic data generation models that power SORDI (Synthetic Object Recognition Dataset for Industries), the largest open-source dataset for industrial AI with over 800,000 photorealistic images across 80 categories. DGX systems deliver an 8x boost in data scientist productivity and 4-6x improved performance over BMW's prior legacy systems; AI training for door-sill defect detection now takes under an hour using as few as five images, and the time required for employees to implement AI automation in QA tasks overall has been cut by more than two-thirds.

The challenge

AI integration in industrial manufacturing confronts challenges in data quality and availability, complex production scenarios, and the need for skilled workforce adaptation. The BMW Group sought to enhance production efficiency by accelerating the speed and cost-effectiveness of AI model training while empowering employees with no-code AI applications.

The solution

The BMW Group leveraged NVIDIA DGX systems to train deep-learning-based synthetic data generation models used to develop the SORDI dataset, the largest and most realistic open-source dataset for the industrial environment. DGX systems were used to implement a complete deep learning operations pipeline, from development and training to deployment, and to build various industrial AI applications such as door-sill and stitch defect detection.

Reported business value

DGX systems delivered an 8X boost in data scientist productivity and 4-6X improved performance over prior legacy systems. Hundreds of thousands of synthetic images are generated at the click of a button, and the time required for employees to implement AI automation in QA tasks has been cut by more than two-thirds.