Updated March 5, 2026
0:00 Welcome to Colaberry AI podcast brought to you by Colaberry AI Research Labs and Carl Foundation. Today, we're doing a deep dive, really getting into the transformative power of generative AI, specifically how it's impacting product engineering and, r and d in the manufacturing sector. Yeah. And that's, especially important right now because, well, the last five years have been incredibly volatile for global markets Uh-huh. Particularly for discrete products. 0:27 Right. You've got shifting customer demands, supply chains getting completely disrupted. Don't remind me. And then this rapid uptake of new tech, like generative AI itself. You know? 0:37 It's created really dynamic, sometimes unpredictable environment. Exactly. And for manufacturers caught in the middle, the core challenge seems pretty fundamental. It's this dual pressure. They need to boost revenue, but also somehow slash costs across the board. 0:53 Absolutely. Across the entire value chain. From the very first design ideas all the way through managing complex global supply networks. And, you know, there's data to back up the importance of engineering here. An IDC report actually found a clear link between higher investment in engineering and r and d and, well, lower cost of goods sold core GS and better revenue growth. 1:14 So strong engineering literally pays off on the bottom line. Precisely. It underlines just how critical product engineering is financially. Okay. So our mission today is really to unpack how generative AI is helping manufacturers tackle these big challenges and unlock those benefits. 1:33 So let's start with the engineers themselves. Their role seems to have changed quite a bit, hasn't it? Oh, absolutely. Fundamentally changed. Product complexity is just well, it's exploded, and everything's so interconnected now. 1:45 You mean, like, software and cars, that kinda thing? Exactly. That's a perfect example. So engineers aren't just mechanical or electrical experts anymore. They really need to be multidisciplinary. 1:53 They're dealing with software, electronics, the traditional stuff all at once. Okay. I get that. And this bigger role must mean they're drowning in data from all sorts of different systems. That's a huge part of it. 2:04 They're constantly interacting with a whole suite of tools and data sources. Like what specifically? Well, you've got product life cycle management. PLM systems, kind of the central hub for product data. Right. 2:15 Then CAD, computer aided design for the digital models, CAM, computer aided manufacturing for production instructions. And if there's software involved, you need ALM, application life cycle management, for requirements. Okay. Plus CAE, computer aided engineering Yeah. For all the virtual simulations and analysis. 2:36 Wow. Okay. That's that's a lot of acronyms, a lot of data streams to manage. It really is. And it sounds like their job isn't just about the core design anymore either. 2:46 The scope's broadened? Definitely. It's much more holistic now right from the start. In the early design phase, they have to think about sustainability Mhmm. And environmental impact, end of life. 2:57 Then there's regulatory compliance, making sure everything meets standards and certifications. Quality has to be baked in from day one. Material choices are critical. And they even need a pretty deep understanding of the supply chain. To make sure they can actually get the parts reliably and affordably. 3:11 Exactly. And with so much more software and physical products, that must be putting a strain on the traditional manufacturing IT folks. You hit the nail on the head. Manufacturing IT might have been more focused on the OT, the operational tech on the factory floor. But now with products being so software intensive, IT needs to support a much wider range of software tools, handle complex data, build out robust infrastructure. 3:35 The pressure is definitely on. Okay. So we've got this picture of engineers dealing with way more complexity, more data, broader responsibilities. Now the exciting part, how does generative AI actually step in to help them manage all this and deliver those key benefits? Let's start with cost reduction. 3:53 Right. Gen AI offers some really significant ways to optimize for cost. Think about design exploration. AI algorithms can explore a vastly larger design space than a human engineer could manually. Okay. 4:04 This means optimizing not just for performance, but specifically for cost effectiveness. Maybe suggesting alternative materials that perform similarly, but cost less. I see. Or identifying design features that are simpler, cheaper to manufacture. It can even factor in sustainability to reduce waste. 4:21 All these optimizations add up to real savings in development and production costs. So it's like finding those clever design tweaks that save money without sacrificing quality. Exactly. We've seen generative design find, say, new structures that cut material use by maybe 18%, but keep the same strength. That's significant. 4:41 Yeah. No kidding. Almost like a cost conscious design partner. What about making better decisions? How does AI help help there? 4:49 Well, think of generative AI as a powerful analysis engine in this context. It can sift through huge amounts of different data types Mhmm. Historical performance data, customer feedback coming in real time, simulation results Mhmm. And find patterns or correlations that humans might miss. It generates these data driven insights. 5:05 Plus, it's great for scenario simulation. You can quickly test what if for different design choices. So you can see the likely outcome before committing. Precisely. It leads to much more informed decisions throughout development, better product quality, products that actually align better with what customers want. 5:20 Like using AI on customer feedback to find needs people haven't even articulated clearly yet. That's a great application. Yeah. Leading to changes that boost satisfaction. So decisions become less gut feel and more data backed. 5:34 Okay. Makes sense. Now you hear a lot about the skills gap in manufacturing. How does Gen AI fit in with productivity and maybe bridging that gap? This is a really critical area. 5:44 For experienced designers, AI tools can automate a lot of the routine time consuming stuff. Freeing them up for harder problems. Exactly. Let's them focus on the complex strategic innovative work. And for the less experienced engineers, generative AI can act almost like a virtual mentor. 6:01 How so? It can offer guidance based on best practices, help them learn faster, avoid common mistakes, and it's not just for new designs. AI can analyze existing designs and suggest optimizations for performance or manufacturability. Or even generate totally new design concepts based on inputs. Yes. 6:18 Based on user defined parameters and constraints, it can generate novel concepts as a starting point. We've seen cases where AI generated initial designs cut assembly time by, say, 12% versus the old way. Interesting. So it helps everyone be more productive, experienced or not. What about just general day to day efficiency for engineering teams? 6:39 The potential there is huge. Engineers spend so much time just finding the right data. It's scattered across PLM, CAD, ERP, everywhere. The data hunt. Yeah. 6:49 Generative AI can offer much smarter search, not just keywords, but understanding the meaning of the query. It can even proactively surface relevant info based on what the engineer is working on right now. Cutting down that admin time. Exactly. Less time searching, more time engineering. 7:03 Plus, natural language interface is just asking the system a question make it even quicker. Right. And that efficiency naturally leads to faster time to market, which is always crucial. Correctly. If you optimize the design process, automate tasks, make decisions faster, the whole development cycle shrinks. 7:19 Getting products out the door quicker to grab market opportunities. That's the competitive edge it provides. Yeah. You could iterate faster, respond to market shifts quicker. Okay. 7:27 And finally, innovation. How does AI actually spark new ideas or solutions? Well, by constantly analyzing all that product data, customer feedback, and learning over time, AI can spot connections or possibilities that might not be obvious to humans. It can explore really unconventional design avenues, evaluate way more options, and suggest genuinely novel approaches. Yeah. 7:52 Fostering more continuous innovation. It might pick up on weak signals and data that point to unmet needs, sparking entirely new product ideas. Like a tireless research assistant always looking for the next breakthrough. Okay. So these benefits are compelling. 8:06 But to actually achieve them, you need a solid foundation for all this engineering data. Right? Has to be secure. Absolutely critical. A secure, well managed data foundation is nonnegotiable. 8:16 We're talking about managing all sorts of data technical specs, products configurations, CAD files, simulation results, process data, a huge mix. And manufacturers already use systems like PLM, ALM, ERP to handle this complexity. Yes. Those are typically the backbone systems. They organize the product info, manage requirements, track resources. 8:37 They form that secure data foundation you need for any digital transformation, especially bringing in generative AI. And crucially, they have security built in to protect IP. And we're already seeing generative AI being plugged into this foundation, especially on the Microsoft cloud. Got any examples? Definitely. 8:53 Take Siemens. They've integrated Microsoft Teams, Azure OpenAI service, and their Teamcenter PLM. What does that enable? It allows real time communication between frontline workers on the factory floor and the engineers back in the office. Faster problem solving, better knowledge sharing right across the life cycle. 9:11 Bridging that gap between design and production. Nice. What about Arris? Arris has brought in AI assist search and an intelligent Copilot using Azure Open AI service and Microsoft Copilot Studio on Azure. It basically changes how users interact with PLM data, makes it much faster to find, analyze, and act on information using scalable search and conversational AI. 9:32 So you can just ask your PLM system a question in plain English, pretty much. Yeah. And get a precise relevant answer quickly. Huge efficiency gain. I can see that. 9:41 And PTC is active here too. Yes. With their CodeBeamer Copilot. It focuses on requirements, authoring, and analysis within their ALM tool. The Volkswagen Group is using it. 9:51 It helps improve efficiency early in the design phase by catching potential requirement issues sooner. Finding problems earlier saves time and money later. Exactly. Especially with complex requirements. And BlueStar PLM. 10:03 How are they using Microsoft AI? They're using Microsoft Copilot for Dynamics three sixty five to automate things. It can automatically generate summaries of engineering objects pulling data from both Dynamics three to two sixty five and BlueStar PLM, and it can auto generate item descriptions in multiple languages. Oh, that's useful. Yeah. 10:21 Makes it much easier to create quotes, BOMMs, invoices for global operations. Okay. So AI is clearly embedding itself into these core data management systems. Now what about accelerating the actual doing of engineering and r and d, the core processes themselves? Right. 10:35 Beyond just managing the data, engineers rely on those sophisticated tools like CAD, CAM, CAE. And that means handling complex data like three d models, manufacturing instructions, huge simulation data sets, plus all the supporting documents and knowledge bases. And generative AI is speeding up these workflows too. Examples. Oh, yeah. 10:54 Look at Harting. They cut the design time for custom electrical connector prototypes from weeks down to minutes. Weeks to minutes. Seriously, how? Using an AI powered assistant built with Azure OpenAI service and Microsoft cloud for manufacturing, it works directly with their Siemens NX CAD system. 11:11 They achieved a 95% reduction in configuration time for these prototypes. That's massive acceleration. Unbelievable. 95%. What about manufacturing programming, the CAM side? 11:21 Hexagon has Proplane AI. It's an AI powered automated CAM programming solution. Part of their cloud platform, Nexus, using Azure OpenAI, Cosmos DB, Azure Databricks, it's reducing machine tool programming time by 75%. Wow. Another huge efficiency jump right on the factory floor. 11:38 Big impact on throughput and turnaround time. And Siemens is embedding AI directly into their own software too. That's right. They have a co pilot for their NXX software. It uses an adapted AI model to understand natural language questions from engineers, give technical insights, streamline design tasks, basically provides AI recommendations and best practices right inside the CAD tool. 11:59 Helping engineers make better choices faster. And ensuring higher quality designs. Okay. Simulation is a big part of r and d. How's AI helping Rescale is doing interesting work here. 12:10 They're integrating AI tools with Microsoft tech to improve simulation data workflows via their Rescale automations platform. It automates data processing, gives real time insights from simulations, improves collaboration. They're using models like five four to cut simulation cycle times and costs while getting more value from the data. Making those complex simulations faster and more insightful. And Siemens made another announcement recently. 12:35 Yeah. Their industrial foundational model or IFM, built on Azure. It's aimed at boosting productivity in engineering and automation tasks, things like automating CAM programming with context aware suggestions, helping generate structured control code for automation, speeding up the creation of PFDs and PNIDs, those crucial engineering diagrams. It really feels like AI is becoming deeply woven into every step. So where does this all lead? 13:00 What's the next big step? I think the next frontier is really the emergence of AI powered digital threads. Digital threads. Explain that. We're moving beyond just optimizing individual tools or tasks. 13:11 We're talking about multi agent AI systems. Think of AI agents that can orchestrate workflows, collaborate across teams, and scale across the entire enterprise, product engineering, supply chain, manufacturing execution, CRM, field service, ERP, all connected. Woah. That sounds ambitious, a fully connected intelligent enterprise. What makes that feasible now? 13:34 Two key things are coming together. First, those unified data foundations we talked about securely pulling and automatically understanding data from all those different systems. Right. Getting the data in one place and making sense of it. Second, the power of generative AI itself to analyze that unified data and actually drive automated actions across the value chain. 13:52 It's the combination that makes the digital thread really possible. So how would these AI agents work in practice? They'd use that rich contextualized data to provide insights or even take action. Imagine an AI agent spotting a potential supply chain problem linked to an early design decision Mhmm. And then automatically suggesting alternative components or suppliers, maybe even initiating the change order process, all proactively. 14:19 That kind of proactive cross functional intelligence. Yeah. That would be a game changer. Are we seeing innovations pushing towards this already? We are. 14:26 RS, for example, introduced Innovator Edge. It's a low code framework for extending these digital thread ecosystems. They're integrating with Microsoft Fabric, m three sixty five Copilot, and the cloud for manufacturing for advanced analytics and AI insights across the thread. And Autodesk, what's their play? Autodesk Fusion is all about connecting people, data, and processes. 14:46 Their data solutions in FusionManage, working with Microsoft Fabric, help with data management and process optimization. And there are other tools, Tandem for digital twins, FlexSim for factory simulation, Fusion Operation for factory ops. They all benefit from better ITOT collaboration, which is key for a digital thread. Makes sense. And FITC is collaborating with Microsoft on this too. 15:09 Yes. They're working on an enterprise data framework and an agentic model for PLM scenarios and wind chill using Microsoft Fabric. The goal is to accelerate digital thread strategies, unlock insights using these AI agents. So lots of partnerships driving this forward, and even manufacturers like Toyota are experimenting with agents. Absolutely. 15:28 Toyota is deploying AI agents in their OBEA or big room system. It's designed to harness collective wisdom and speed up innovation. They have agents like a vibration agent or a fuel consumption agent bringing experts together virtually to solve problems faster. Fascinating. Using AI to break down silos and accelerate innovation. 15:46 So pulling it all together, integrating these AI solutions and agents, what's the bottom line for manufacturers? Well, it unlocks significant innovation potential, leads to real cost reductions, and boosts operational efficiencies dramatically. Ultimately, it makes manufacturers much better equipped to handle today's challenges and, importantly, to seize new opportunities. It's about building resilience and a competitive edge. A truly profound shift for the industry. 16:13 If listeners wanna learn more, where should they go? A few good places. Definitely check out the Microsoft Cloud for Manufacturing website. Also, Microsoft for automotive, if that's their specific sector. Mhmm. 16:23 And the Microsoft AI and action site have lots of case studies and examples. Good resources. And any chance to see this stuff live? Keep an eye out for Hanover Mes twenty twenty five. Microsoft and many partners will be there showcasing these kinds of solutions for product engineering and r and d. 16:37 It's a great place to see it in action. Excellent tip. Well, thank you. This has been an incredibly insightful deep dive into generative AI's impact on manufacturing engineering and r and d. D. 16:48 Really fascinating stuff. My pleasure. It's definitely an exciting space to watch constantly evolving. Thank you for listening in. Subscribe and follow Colaberry on social media links in the description, and check out our website, www.colaberry.ai backslash podcast for more insights like this.
