“Our aim is to reduce the time spent on technical drawings by up to 60% and the time spent on documentation by up to 50%, while ensuring accuracy and consistency.” C. K. How can engineering and artificial intelligence work together to accelerate innovation? “Current AI models produce approximate results that work well when writing texts or creating art, but don’t work in engineering.” H. G. CHRISTIAN KOPP Senior Executive Vice President, Director of the Exterior & Lighting Business Group Design for Assembly (DFA) validation, reduce the time spent on documentation by up to 50%, and improve the accuracy and consistency of all deliverables. What is the strategic value of this first stage for OPmobility? C. K. By integrating AI tools into our existing systems, via a platform designed for our industry, we are going to accelerate our workflows by automating tasks related to technical drawings and the creation of manuals and catalogs. This will also have a direct impact on the different stages of production, making it possible to provide our industrial sites with more efficient assembly instructions. We are a global group and it is vital that we can quickly distribute content in multiple languages across our network. In a market environment in which speed is a key factor, AI and Foundation EGI’s platform provide considerable potential value. For example, Foundation EGI has enabled an automotive manufacturer to reduce the time spent on technical drawings by 60%, thereby accelerating the DFA validation process. THE SPONSOR — — THE EXPERT HARSHIT GUPTA Head of Business Development & Operations at Foundation EGI How did you overcome this difficulty? H. G. Foundation EGI is the result of 10 years of research conducted at MIT. We believe that, at its core, engineering is programming. It’s a matter of logic. And all logical systems can be coded in software. Consequently, we’re developing specific languages to translate engineering data and processes into software. Our platform is designed to accelerate every stage of the product innovation process, from design to documentation, including exploratory design, sourcing, simulation, and production. Can you give us some specific examples of this acceleration? H. G. Sometimes, designers produce specifications that aren’t realistic and will result in multiple iterations between design and production teams. Using a 3D model, the EGI platform will help them generate the right specifications from the outset. Assembly instructions are another example: AI will be able to produce them from 3D models. With sequences calculated to perfection, assembly lines will become faster. Can you tell us a bit about Foundation EGI? Harshit Gupta. We are a start-up that emerged from MIT (the Massachusetts Institute of Technology). We aim to redefine how engineering teams work in the real world. Automating documentation and design is a way to accelerate innovation. It feels like AI can already do everything we need. Why was it necessary to create this platform? H. G. Big Tech has spent significant sums of money on building what is known as Artificial General Intelligence (AGI). But it doesn’t work when dealing with engineering processes in the real world. For example, if you ask AI to create an exploded view of an automotive parts assembly diagram, it will produce something that looks like a child’s drawing. There are several reasons for this: AI doesn’t understand geometry in space, lacks engineering expertise, and doesn’t have a good grasp of the data required for critical engineering. Most AI models are based on the principle of neural networks. They process enormous volumes of data to generate results. This process, which is essentially statistical, produces approximate results that work well when writing texts or creating art, but don’t work in engineering, something that demands accuracy and attention to detail. Moreover, these models are “black boxes”: you don’t know why AI has generated a particular result. The results aren’t reliable for use in engineering! Scan the QR code to watch the MIT Symposium. OPEN MIND What are the challenges facing OPmobility’s engineering teams? Christian Kopp. Today, our development processes are fragmented. Our teams use several tools, including CAD (ComputerAided Design), PLM (Product Lifecycle Management), and various documentation systems. This manual approach to creating technical content slows down our products’ time to market. On top of this, there’s the complexity of providing this documentation in all the languages used within the Group. We need to make this content quickly accessible to all our teams around the world. What does artificial intelligence offer in response to these challenges? C. K. AI can reduce the time spent on technical documentation by approximately 50%, while improving the accuracy and consistency of deliverables. It can also be securely integrated into our existing CAD and PLM systems, ensuring a seamless workflow. What issues will you be working on with Foundation EGI? C. K. Our aim is to develop an AI stack for the Lighting Business Group that can automate key engineering processes and generate catalogs of 2D and 3D diagrams directly from CAD files, assembly instructions, technical drawings, and multilingual manuals. This approach will reduce the time needed to produce technical drawings by roughly 60%, accelerate 28 • • 29
RkJQdWJsaXNoZXIy NTkwMzk=