Best of LinkedIn: Smart Manufacturing CW 39/ 40

00:00:00: This episode is provided by Thomas Allgaier and Frennus based on the most relevant LinkedIn posts about smart manufacturing in calendar weeks, thirty nine and forty.

00:00:09: Frennus is a B to B market research company that supports enterprises across the smart manufacturing industry with a market, customer and competitive insights.

00:00:18: they need to navigate dynamic markets and drive customer centric product development.

00:00:23: Welcome to the deep dive.

00:00:26: Today, we're really trying to cut through some of the noise that you see online around smart manufacturing.

00:00:31: For professionals in the field, I think the challenge isn't finding information right.

00:00:35: It's finding the actionable stuff.

00:00:36: Exactly.

00:00:37: What can you actually use?

00:00:39: Our mission here is to pull out the top insights, the strategic prerequisites, and maybe the core tech trends from the last couple of weeks of industry conversation on LinkedIn.

00:00:48: Yeah, and what really struck me this time was a clear consensus, you know.

00:00:51: People seem kind of done talking about the hypothetical smart factory.

00:00:55: The focus now is really on foundational readiness.

00:00:57: Foundational

00:00:58: readiness.

00:00:59: Meaning data quality, cultural alignment, and... really smart strategic deployment.

00:01:06: We saw lots of talk, not just about robots and AI, the flashy stuff, but about the systems connecting at all, MES, five G infrastructure.

00:01:14: It's about looking at what's actually working on the ground.

00:01:18: Okay,

00:01:18: that makes sense.

00:01:19: Let's dig into that then with a theme that just kept popping up.

00:01:22: Building the foundation.

00:01:24: It's maybe not the most exciting part of industry.

00:01:26: four point zero, but based on what we saw.

00:01:29: Trying to skip it.

00:01:30: That seems like guaranteeing failure.

00:01:32: Oh, absolutely.

00:01:32: It's the prerequisite for pretty much everything else.

00:01:35: Jeff Winter put it really well, actually.

00:01:36: He used that Maslow's hierarchy analogy.

00:01:39: Oh,

00:01:39: yeah.

00:01:39: for factories.

00:01:40: Yeah, basically saying you need basic stability and uptime.

00:01:42: First, kind of the physiological needs of the plant before you can even think about innovating.

00:01:47: You just can't cheat that climb.

00:01:48: If your basic processes are unstable, fancy tech just means you break down faster.

00:01:53: Right, faster breakdowns, not the goal.

00:01:55: So what does that readiness look like practically?

00:01:57: Sven Diedrich, he outlined three things he called non-negotiable prerequisites.

00:02:03: He argued that successful automation, it doesn't start with buying a robot.

00:02:07: It starts way earlier.

00:02:09: with getting your house in order.

00:02:11: Things like consistently high quality master data.

00:02:15: Super important.

00:02:15: Like product weight, dimensions, the basics.

00:02:18: Exactly.

00:02:19: And then system-based production planning and importantly having a robust manufacturing execution system and MES already up and running.

00:02:28: That data quality piece is just everything, isn't it?

00:02:30: Without it, you're just... Well, automating garbage, as they say.

00:02:34: And Chris Sturgy really hammered on the planning stage before you automate anything.

00:02:38: What did he emphasize?

00:02:39: He stressed, you have to be ruthless in answering three questions first.

00:02:42: Why are we doing this?

00:02:43: Critically important.

00:02:44: Then, what specifically are we automating?

00:02:47: And only then, how are we automating it?

00:02:49: Skipping that?

00:02:50: why, apparently that's the root cause of so many pilots just fizzling out.

00:02:53: Hmm,

00:02:54: the why.

00:02:54: I think, though, the toughest part of that foundation might not even be the data or the tech.

00:03:00: It's the people.

00:03:01: the culture.

00:03:02: Always the hardest part.

00:03:03: Jacobo Luré-Casal had this really blunt warning.

00:03:06: He said, AI is not some magic wand for a broken factory.

00:03:10: If your processes are messy, data is dirty, and, crucially, your workforce doesn't trust the systems.

00:03:17: Forget it.

00:03:17: You won't see results.

00:03:18: It's a culture challenge first.

00:03:19: It absolutely is.

00:03:21: And building that culture requires real internal partnership.

00:03:25: Rory to Go highlighted this successful AI adoption needs alignment right across the board.

00:03:30: You need IT, C-level, and operations leaders all pulling in the same direction.

00:03:34: That makes sense.

00:03:35: If Ops doesn't trust the IT platform, or maybe IT doesn't get the shot for realities.

00:03:39: Yeah, the project's basically dead on arrival.

00:03:41: It just won't

00:03:42: work.

00:03:42: But the upside, if you do solve that people and process piece.

00:03:46: It looks enormous.

00:03:47: Jack Truman shared observations showing that systems really designed for the frontline user, easy to use, real-time feedback, mobile, maybe they drive massive gains, and quickly.

00:03:56: Out

00:03:56: quickly.

00:03:57: He's talking like up to twenty-six percent productivity boosts in just ninety days, often translating into serious savings, like millions annually.

00:04:05: The transformation really starts on the shop floor.

00:04:07: That actually leads perfectly into our next theme, because the MES is often that key system connecting the frontline user right back to the enterprise data flow.

00:04:15: Right, the MES, Muhammad al-S.A.

00:04:16: It's stressed.

00:04:17: It's not just some digital ledger.

00:04:19: He called it the central nervous system for efficiency, quality, traceability.

00:04:24: It's

00:04:24: the critical link moving manufacturers towards that, you know, true smart factory vision.

00:04:29: It bridges the physical shop floor with the sort of abstract world of enterprise systems.

00:04:35: Exactly.

00:04:36: And here's where we saw some really interesting innovation kind of.

00:04:39: challenging the traditional vendor model.

00:04:42: Mahesh Adnani shared this amazing case study.

00:04:45: His team got quoted, wait for it, twenty five million dollars for a traditional MES implementation.

00:04:51: Wow.

00:04:52: Instead of paying that, they built their own solution in-house, apparently superior user first, with just three developers.

00:04:58: Hold on.

00:04:58: They built a better MES themselves.

00:05:01: for less than twenty five million dollars with three people.

00:05:03: That's

00:05:03: the story.

00:05:04: Yeah.

00:05:04: Astonishing.

00:05:05: How worried should the big enterprise software players be about that?

00:05:09: That sounds like, well, a potential shift, right?

00:05:12: Power moving to internal teams who really know the manufacturing process inside out.

00:05:16: Well, it definitely signals a desire for solutions built for manufacturing engineers, not just for the IT department's checklist, maybe.

00:05:24: And that focus on internal capability, it leads us straight to the next level.

00:05:28: AI.

00:05:28: Right.

00:05:29: Artificial intelligence.

00:05:30: Mila Wordzowski talked about the necessary evolution here, moving from, you know, static automation systems, just repeating a fixed task to what he called adaptive intelligence.

00:05:40: Systems that can learn, maybe predict the root cause of a problem and even suggest corrective actions autonomously.

00:05:46: That adaptive capability, that feels like the real promise of industry.

00:05:49: four point zero.

00:05:51: Busaratu saying summarize the pillars that unlock this potential.

00:05:55: condition monitoring, predictive maintenance, digital twin tech, and automated quality control.

00:06:00: A

00:06:00: big four.

00:06:01: Yeah.

00:06:01: And when they're implemented together, holistically, the returns are supposedly massive, like forty, fifty percent downtime reduction, ten, twenty percent efficiency gains, big numbers.

00:06:11: But, and there's always a but, all that intelligence, it relies heavily on connectivity.

00:06:17: And probably not just the old Wi-Fi network struggling in the corner.

00:06:20: Definitely

00:06:20: not.

00:06:21: Louis C stressed this AI connectivity imperative.

00:06:24: Think about the sheer volume of data we're talking about.

00:06:26: Some factories are already generating five terabytes of data daily.

00:06:30: Five

00:06:31: terabytes a day.

00:06:32: Yep.

00:06:32: And that's projected to quadruple by twenty twenty six.

00:06:35: It's like, I don't know, downloading hundreds of HD movies every single hour just from machine sensors.

00:06:41: Yeah.

00:06:42: Yeah.

00:06:42: Trying to push that through legacy infrastructure.

00:06:45: impossible.

00:06:45: Totally impossible.

00:06:46: And Erickson's research backs this up.

00:06:49: They found eighty eight percent of US businesses now see five G as absolutely critical for optimizing AI at that kind of scale.

00:06:57: It's becoming the necessary plumbing.

00:06:59: And we're actually seeing real-world large-scale deployments proving this now.

00:07:02: Andres Torres highlighted Hitachi's new U.S.

00:07:05: railcar factory.

00:07:06: It's running its operations, digital twins, autonomous vehicles, even those robot dogs for defect detection.

00:07:11: The Boston Dynamics ones?

00:07:13: Possibly, those types, yeah.

00:07:15: All running on a private, five-G network from Ericsson and Globalogic.

00:07:19: And Peter Linder added that this isn't just happening in the U.S Airbus is also deploying private five-G across strategic global sites.

00:07:26: So the infrastructure investment... is happening.

00:07:29: Okay, so from connectivity, let's move to the physical layer.

00:07:33: Robotics.

00:07:34: This is maybe the most visible part of smart manufacturing and the conversation seems to have moved way beyond just replacing jobs.

00:07:40: Exactly.

00:07:41: Tom Scooball have framed it really well.

00:07:43: It's about man with machine.

00:07:45: The real goal of automation isn't just replacement, it's removing the repetitive inefficient tasks.

00:07:50: This frees up skilled people to do higher value complex work and the difference in capability.

00:07:56: It's almost funny.

00:07:57: Michael Hofacker pointed out that in sheet metal fabrication, for instance, an industrial robot reacts in under ten milliseconds.

00:08:03: A human.

00:08:04: Around two hundred milliseconds.

00:08:05: That level of precision and speed.

00:08:07: Humans just can't match it for certain tasks.

00:08:10: Right.

00:08:10: But the big hurdle for robotics used to be programming, didn't it?

00:08:13: Complex, slow, needed specialists.

00:08:15: Expensive.

00:08:17: It was, but that barrier seems to be really coming down.

00:08:19: Dario Stodacic announced ABB's new AI coach, built right into their robot studio software.

00:08:26: It uses generative AI to simplify and speed up programming, making commissioning faster, less specialized.

00:08:32: Interesting, so AI help and program the robots.

00:08:35: Yeah, and we saw something similar from Siemens too.

00:08:38: Roland Locker showed how BSH Home Appliances is using Siemens Process Simulate Co-Pilot.

00:08:44: It apparently cuts complex optimization times for robotic paths from hours down to just minutes.

00:08:50: Again, empowering engineers who maybe aren't deep robotics experts.

00:08:53: And the innovation looks like it's getting genuinely disruptive now.

00:08:57: Robert Little detailed this collaboration between McKenna Labs and Toyota.

00:09:01: They're working on something called roboforming.

00:09:02: Roboforming,

00:09:03: what's that?

00:09:04: robots and AI to actually shape metal body panels without needing the huge, heavy, fixed tooling, those massive, twenty ton dies that have been standard and automotive for, well, forever.

00:09:14: Whoa, shaping panels without fixed dies, that's not just innovation.

00:09:18: That potentially changes the entire workflow of fabrication, doesn't it?

00:09:22: Seems like it.

00:09:23: If you can rapidly shape a panel without tooling, you could unlock true mass customization, much faster changeovers.

00:09:30: It feels like a massive potential chef.

00:09:32: Huge implications.

00:09:33: And that sense of major market upheaval was kind of underlined by that proposed big acquisition.

00:09:39: SoftBank looking to buy ABB's robotics division for what was a five point four billion dollars.

00:09:44: Mustafa Mohammed Sayed flagged that one.

00:09:46: Yeah, that's a serious move.

00:09:48: It signals a clear strategic bet on what people are starting to call physical AI.

00:09:53: Physical AI.

00:09:53: Robots that can think, adapt, collaborate, learn autonomously right there on the factory floor, not just programmed automatons.

00:10:02: Okay.

00:10:03: So, intelligent, adaptable robots.

00:10:05: All of this, though, it needs significant investment.

00:10:07: Of course.

00:10:08: Big capital flows.

00:10:09: Jamie Codham noted a major commitment of capital into U.K.

00:10:12: manufacturing specifically, including a five hundred million government boost focused on innovation and efficiency.

00:10:18: And for companies in Europe, getting that capital strategy right seems key.

00:10:22: I want our Chish Naska-Hopane gave some valuable advice.

00:10:25: to really maximize European manufacturing grants which can apparently cover up to seventy percent of a project.

00:10:30: Seventy percent, that's huge.

00:10:32: Right, but the key is you need to integrate your location strategy and your financial planning right from the very beginning.

00:10:38: Don't finish the tech plan and then think about funding.

00:10:40: It has to be integrated early.

00:10:42: Smart advice.

00:10:44: Okay, so bringing this all together.

00:10:47: The tech is clearly advancing rapidly.

00:10:49: The capital seems to be flowing, but we probably need a bit of a reality check

00:10:53: before we wrap up.

00:10:54: Yes, absolutely.

00:10:55: The enthusiasm needs tempering with where companies actually are.

00:10:59: Sebastian Schmitz's report from that rethink, smart manufacturing event, it confirmed that despite all the hype, most companies are still pretty far from, let's say, holistic AI application.

00:11:11: Still early days in practice for many.

00:11:13: Yeah.

00:11:14: that shop floor data we talked about all those terabytes.

00:11:16: It's often just ending up on dashboards.

00:11:18: For monitoring, it's not necessarily being fed back in operational systems for genuine predictive use.

00:11:24: The data loop isn't closed, the data is kind of trapped.

00:11:26: Which circles right back to that complexity and integration challenge?

00:11:30: Michael Donahue's analysis of the big challenges for twenty twenty five, twenty twenty six, things like cost escalation, maybe weaker demand, legacy tech debt, talent gaps.

00:11:40: The usual suspects.

00:11:41: Pretty much.

00:11:42: His point was that.

00:11:43: Manufacturers might need to shift focus slightly, prioritize resilience perhaps over pure efficiency in the near term.

00:11:49: And the path forward involves integrating external partners really smartly and crucially, achieving that full convergence between IT and OT systems.

00:11:58: Ah, the IT OT convergence.

00:12:00: Gee, Vikram, champion that idea.

00:12:02: It's kind of the holy grail, isn't it?

00:12:03: Getting that seamless data flow across the entire enterprise.

00:12:06: But it's still hard.

00:12:07: Very hard.

00:12:08: IT and OT often operate on different protocols, different management structures, sometimes completely different cultures even.

00:12:14: But without bridging that gap, all the intelligence we've discussed stays siloed.

00:12:18: It can't unlock its full value.

00:12:20: Okay, so we've seen the tech is ready, capital is flowing, the market's definitely shifting towards adaptive intelligence, physical AI.

00:12:27: But here's a final thought.

00:12:29: If talent gaps and change fatigue are still listed as major operational risks, how can manufacturers really ensure that cultural adoption, that leadership readiness for these powerful AI tools before they make the huge capital investment?

00:12:46: That cultural piece feels like it needs to be front and center in the planning.

00:12:49: Maybe it's the most important piece of

00:12:58: the puzzle.