How AI Is Accelerating Engineering Simulations

In Episode 16 of Espresso Talks, we explore how Physical AI is changing the way engineering teams approach simulation, design, and data.

Unlike traditional AI models that learn patterns from data alone, Physical AI combines data-driven approaches with physical laws and engineering knowledge. This opens the door to using AI as a faster, more efficient alternative for computationally expensive simulations—making it possible to explore more design variations and accelerate engineering workflows.

But bringing AI into engineering is about more than speed. The discussion also looks at some of the key challenges behind real-world adoption: how to build trust in AI-generated results, make models interpretable, prevent physics-based “hallucinations,” and ensure that valuable engineering data and intellectual property remain under a company’s control.

As engineering workflows evolve, AI can take on more of the repetitive work while engineers remain at the center of critical decisions, applying their expertise and intuition where it matters most.

In this episode, Jamarco Gabriel, AI Team Lead at Sphere Energy, and Oliver Schilter, Lead AI Architect, discuss where Physical AI stands today, how companies can prepare their engineering data for AI, and what the future of AI-assisted engineering could look like.