In a commendable achievement for both Ignition Computing and its technology partner CelSian Glass & Solar, their joint project proposal titled “Quality Predictions and Simulation Optimisation for Glass Production” has been recognized as the top-ranking submission in the latest MIT AI R&D grant program.
This accolade comes from the Netherlands Enterprise Agency (RVO), which oversaw a competitive selection process involving 57 applications.
The SME Innovation Stimulus for Regional and Top Sectors, abbreviated as MIT, serves as a pivotal initiative aimed at fostering creativity and innovation among Small and Medium-sized Enterprises (SMEs) in the Netherlands. This program is dedicated to recognizing and funding projects that exhibit extraordinary potential in addressing global challenges while simultaneously enhancing the nation’s economic landscape. The MIT R&D AI program specifically emphasizes collaborative projects that harness artificial intelligence to drive innovations across the top sectors.
The evaluation criteria for the grant emphasized objectivity, focusing on a proposal’s potential impact, innovativeness, sustainability, and feasibility. Ignition Computing’s project achieved an impressive score of 94.9 out of 100 points, underscoring its viability and the significant promise it holds for the glass production sector.
CelSian and Ignition Computing Secure MIT R&D AI Grant for Glass Production Innovations
With this grant funding, Ignition Computing aims to develop advanced simulation acceleration tools tailored for complex multi-physics problems inherent in glass manufacturing. The integration of these tools into the production processes will likely revolutionize the industry, contributing to enhanced efficiency and quality.
CelSian brings decades of expertise in computational fluid dynamics (CFD) tools, particularly in glass quality prediction. The collaboration will see Ignition Computing adapt and enhance its machine learning-based tool known as Preconnet, in conjunction with CelSian’s renowned GTM-X tool. Preconnet offers a promising framework for accelerating simulations by generating preconditioners, establishing initial guesses, and suggesting optimal solver configurations. Originally developed for plasma simulations in collaboration with Plasimo, Preconnet’s versatility allows it to address a range of new simulation challenges, including those facing the glass industry.
The glass melting process, which can extend over several days, presents substantial challenges, primarily stemming from temperature and material fluctuations. These fluctuations can lead to detrimental product quality issues, including the formation of bubbles within the glass—sometimes resulting in significant production failures and batch losses. As a response, the integration of machine learning and simulation technologies through this project is poised to yield more precise predictions of glass quality, ensuring a marked improvement in production efficiency.
The anticipated benefits from this initiative are manifold, with expectations of enhanced quality prediction models that lead to significant economic and environmental improvements. By optimizing feedstock utilization and reducing energy consumption, each glass manufacturing furnace could potentially achieve annual energy cost savings ranging from EUR 250,000 to EUR 1,250,000, along with appreciable reductions in CO2 and nitrous oxide emissions. Such positive impacts resonate deeply with our shared vision of fostering a sustainable industrial future.
The collaboration between Ignition Computing and CelSian, bolstered by the MIT R&D AI grant, represents a promising advancement in the glass manufacturing sector. As the project unfolds, it holds the potential to redefine standards of quality, efficiency, and sustainability within the industry, firmly establishing a path toward the innovative and environmentally-conscious manufacturing practices of tomorrow.
Source:Ignition Computing with additional information added by GlassBalkan
