The Omniverse platform, a virtual factory planning tool, integrates planning data and applications via a data layer in the cloud, enabling real-time collaboration without compatibility limits. Omniverse brings together data from multiple professional design and planning tools from different manufacturers to generate photorealistic real-time simulations in a single collaborative environment.
“The BMW Group has mastered high-volume production of individualised automobiles – their business processes are among the most complex in the world,” said Jensen Huang, founder and CEO of Nvidia, at his company’s GTC developer conference. “In their vision of the factory of the future, humans and robots work together, engineers from all disciplines of factory design collaborate in a shared virtual space. The entire factory is simulated with the highest photorealistic precision.”
Virtual factory design is by no means new. Until now, however, data from different applications had to be transferred, which is not only time-consuming, but also leads to compatibility difficulties and does not always take into account the most up-to-date data. Omniverse is supposed to offer the possibility to combine live data from all relevant databases in a common simulation. Re-importing data will thus become superfluous. The aim is to assess changes and adjustments at an early planning stage and to generate an overall view. This high level of transparency enables planners and production specialists to plan highly complex production systems more accurately and quickly without interface losses and compatibility problems.
Photorealism meets data coherence
An important advantage of Omniverse is that employees at different locations in different time zones have access to the virtual simulation at any time and can jointly plan or optimise a process or production system in detail. The areas of application are numerous: in future, planners can work with production specialists on real-time data that is synchronised in the Omniverse cloud infrastructure. But the integration of new plants can also be discussed with suppliers. Structural and facility data should be just as integrable as positions and part numbers of material in the production process. The platform can also be used for logistics planning and enables a previously unknown data continuity from planning to production.
The ability to make immediate live changes will noticeably increase the speed of decision-making processes in the future. Accelerated by scalable GPU performance, BMW Group production planners will be able to visualise the entire planning lifecycle for each plant in the global production network. This will be supported by a wide range of AI-enabled use cases – from autonomous robotics to predictive maintenance and data analytics.
BMW Group and Nvidia have been working together for some time. In an initial pilot project, for example, the US company equipped logistics robots as well as autonomous transport robots (STR) developed in-house with high-performance technology and special AI modules for control.