Real-Time Digital Twins Using CFD, Sensors & AI - Haris Kokkinos | Podcast #165

About this video

• Video Title: Real-Time Digital Twins Using CFD, Sensors & AI - Haris Kokkinos | Podcast #165
• Channel: Jousef Murad | Deep Dive
• Speakers: Jousef Murad, Haris Kokkinos
• Duration: 21:17

Overview

This podcast episode features Haris Kokkinos, who discusses the concept and application of real-time digital twins. He explains how digital twins are created by combining physics-based simulations with sensor data and AI, and how this technology is used in various industries, including defense. The conversation also touches upon the evolution of engineering roles, the importance of integrated software tools, and the future of AI in simulations.

Key takeaways

• Defining Digital Twins: Digital twins are not a single technology but a combination of physics-based simulations, sensor data, and AI, used to create a dynamic virtual representation of a physical asset.
• Building a Digital Twin: The process involves initial simulations (like CFD and structural analysis), integrating real-world data from sensors, and applying AI algorithms (such as reduced-order models) to achieve real-time performance.
• Challenges in Digital Twins: A significant challenge is training AI to predict infinite real-world scenarios, especially when physical testing data is limited early in a project.
• Role of Physical Testing: Physical testing is crucial for validating simulation results, especially once the product is manufactured.
• Software Integration: Modern engineering relies on integrated software platforms (like Siemens' Simcenter suite) that allow for easy data exchange between different simulation and hardware components.
• Evolution of Engineering: The role of engineers is changing, requiring new skill sets in areas like AI. It's becoming less common for a single engineer to be an expert in both traditional engineering fields (like CFD) and AI, necessitating team specialization.
• Learning FEA: Engineers can learn Finite Element Analysis (FEA) by practicing extensively, understanding the underlying theory, and using simulation software like Simcenter 3D, while acknowledging that failure is a necessary part of the learning process.
• Hiring Engineers: When hiring, companies look beyond just degrees and prior experience, considering personality, willingness to collaborate across disciplines, and a proactive approach to problem-solving.