Case Study
Case Study
“Our solution integrates AI-based high-end fluid simulation into the digital twin of the filling machine, opening up completely new possibilities for forecasting, optimization, and visualization.”
— Thomas Albrecht
Head of Simulations, Krones
Krones, a global process manufacturing line provider for bottling, canning, and packaging, encountered a limiting obstacle: time-consuming fluid simulation methods developed decades ago.
Liquid dynamics in manufacturing are inherently fast-paced and unpredictable, requiring precise management at every stage. Velocity, design, and flow dynamics create significant challenges for controlling and simulating uncapped bottles, as every variable — including pressure, flow, and turbulence — must be meticulously controlled and predicted.
The inability to optimize these factors in real time can lead to excessive waste, inefficiencies, and operational problems during subsequent phases, making advanced simulation technologies essential for maintaining efficiency and minimizing losses.
Evaluating liquid dynamics in bottling and filling lines demanded one to two hours per simulation, restricting both innovation and the speed of operational upgrades. Overcoming this pressure point became more urgent as market expectations shifted toward digital, data-driven manufacturing.
Realizing the immediate need to transition from legacy workflows to comprehensive digital processes, Krones set clear priorities: cut simulation times, empower real-time decision making, and double down on the company's commitment to advanced industry practices.
This stack provided exactly what Krones needed: reliability, speed, scalability, and a demonstration platform to drive a strategy in line with Industry 4.0.
Together with Ansys, part of Synopsys, Krones revolutionized its fluid simulations with CADFEM, Microsoft, NVIDIA, and SoftServe.
Ansys Fluent software and NVIDIA Omniverse libraries brought the two worlds of accurate computational fluid dynamics (CFD) and virtual assembly lines together. Fluent native GPU solver simulations were deployed seamlessly into the Krones Microsoft Azure account to provide access to High Performance Computing (HPC) infrastructure to improve simulation time and data management. SoftServe helped the partners implement these solutions on Azure, and CADFEM created models that form a physically accurate digital twin and supported the customer directly in the Fluent simulations (run on NVIDIA H100 GPU).
Krones built a physically accurate agentic digital twin by harnessing the combined expertise of these strategic partners, enabling operational agility and faster innovation.
如果您面臨工程挑戰,我們的團隊將隨時為您提供協助。憑藉豐富的經驗和對創新的承諾,我們邀請您與我們聯絡。讓我們共同合作,將您的工程障礙轉化為成長和成功的機會。立即與我們聯絡,開始對話。