Best of LinkedIn: Digital Construction CW 38/ 39

00:00:00: This deep dive is provided by Thomas Algeyer and Frennus based on the most relevant LinkedIn posts about digital construction in calendar weeks thirty-eight and thirty-nine.

00:00:09: Frennus is a B to B market research company that supports enterprises across the construction industry with the market, customer, and competitive insights they need to navigate dynamic markets and drive customer-centric product development.

00:00:23: And that context is really important, actually, because looking at the posts from those two weeks, you definitely see a shift happening in digital construction talk.

00:00:31: Oh, yeah.

00:00:32: How so?

00:00:33: Well, it feels like we're moving past just saying digitization generally and really starting to focus hard on, you know, specific functional value.

00:00:41: Getting specific.

00:00:42: Exactly.

00:00:43: It's less about just adopting tech for tech's sake and more about tackling those really painful, costly problems everyone faces on site or in the office.

00:00:51: Right.

00:00:52: So our mission today is to take a deep dive into those top trends, and importantly, the actionable insights from LinkedIn during that period.

00:00:59: We've sort of grouped them into four main areas.

00:01:01: First, AI and data strategy that's huge right now.

00:01:05: Then the ongoing evolution of BIM and standards.

00:01:08: Third, Forty visualization.

00:01:11: And finally, the sometimes tricky reality of site automation.

00:01:15: Makes sense.

00:01:16: Yeah.

00:01:16: And if we're talking tech, well, you almost have to start with data, don't you?

00:01:19: Absolutely.

00:01:20: Okay, theme one.

00:01:21: AI and data strategy.

00:01:23: So AI, everyone's talking about it.

00:01:25: And the general feeling, like Danny McCready put it quite well, seems to be it's not about replacing people.

00:01:30: Right, augmentation.

00:01:31: Yeah, augmenting them, freeing up project managers from, you know, endless admin to focus on the human stuff, the high value work.

00:01:39: That sounds great.

00:01:40: It does.

00:01:41: But there's always a but.

00:01:42: What's the biggest hurdle people are actually hitting when trying to get value out of AI?

00:01:46: Is it cost?

00:01:47: Complex.

00:01:48: But interestingly, what keeps coming up isn't the AI tech cost itself.

00:01:51: It's the data.

00:01:53: Ah, the data foundation.

00:01:54: Exactly.

00:01:55: Jeff Daly pointed this out.

00:01:56: Many companies just don't have a solid data strategy.

00:02:00: And Anani Arab has some fascinating analysis from Australia.

00:02:03: Yeah, saying even with AI adoption rising, success really hangs on getting your data sorted first, consolidating it, classifying it, governing it.

00:02:14: Because, get this, the median construction firm operates across eleven disconnected data environments.

00:02:20: Eleven, seriously?

00:02:21: Eleven, so think about it.

00:02:23: How can an AI possibly work effectively if it's wading through, like, a decade of siloed spreadsheets in different CDEs?

00:02:31: The input is chaos.

00:02:33: Garbage in, garbage out.

00:02:34: Okay, so that highlights the skills gap too, right?

00:02:37: Christiansen warned about teams falling behind without proper training.

00:02:40: It's not just plug and play.

00:02:41: Not at all.

00:02:42: And how critical is domain expertise for the AI tools themselves?

00:02:46: Hugely critical.

00:02:47: Right.

00:02:47: Ian Yeo explained this brilliantly.

00:02:49: Generic AI just fails.

00:02:50: Yeah.

00:02:50: Because it doesn't speak construction.

00:02:52: It doesn't get the jargon.

00:02:53: Exactly.

00:02:54: It needs tools that understand what a kicker is, you know, the concrete upstand or what striking form work actually involves.

00:03:00: The AI doesn't get that context.

00:03:02: It's not helpful.

00:03:03: It's potentially dangerous.

00:03:04: That makes perfect sense.

00:03:05: And it flags a bigger structural issue Guido Machiochi brought up.

00:03:09: He argued transformation stalls because you have this mismatch.

00:03:12: What kind of mismatch?

00:03:13: Well, traditional consultants know the theory, but maybe not the site reality, while internal IT might lack the really deep advanced

00:03:21: tech skills.

00:03:23: Yeah, I see that.

00:03:24: So Guido's point was you need these AEC native tech partners, people who have actually lived the challenges on site and understand the tech deeply.

00:03:32: That bridge is vital.

00:03:33: and it leaves right into a strategic point Omri Stern made, leaders face a choice.

00:03:38: Build AI in-house or partner up.

00:03:41: Build versus buy, essentially.

00:03:43: Kind of.

00:03:43: Building in-house, like Caterpillar does for its predictive maintenance, gives you total control over proprietary data.

00:03:50: Big advantage.

00:03:51: But

00:03:51: slow, maybe?

00:03:52: Potentially.

00:03:53: Partnering with platforms, say Procore or Joan Software, gets you moving faster.

00:03:58: But it's a big decision.

00:03:59: Omri's Dressed, you need a proper framework, analyze the value, the cost, the risks.

00:04:04: Right, it needs structure.

00:04:05: Okay, so if data strategy is that foundation, BIM is kind of the system trying to organize it all, right?

00:04:11: Let's shift to theme two.

00:04:12: Yeah, absolutely.

00:04:13: And we have to clear up that common confusion first, the one Nikolayovitch mentioned.

00:04:18: Revit is just a tool.

00:04:19: Right.

00:04:19: Not the whole thing.

00:04:20: No.

00:04:20: BIM is the methodology.

00:04:22: It's the whole system process, collaboration, managing data.

00:04:25: It involves multiple tools like Navisworks, CDEs, all working together.

00:04:30: That's such a key distinction.

00:04:31: So how do we get past just making nice looking, three-D models to actually getting, you know, real project value out of BIM?

00:04:40: Well, Sumana and Dorothy really hit on this.

00:04:42: You have to begin with the end in mind.

00:04:44: To find the purpose up front.

00:04:45: Exactly.

00:04:47: Before you even start modeling, what is this BIM model for?

00:04:50: Is it purely for clash detection during design?

00:04:54: Or is it for facilities management years down the line without that clear goal?

00:04:58: It just becomes an expensive picture.

00:05:00: And that clarity often gets lost when different teams use different software, right?

00:05:05: Which leads to what Conrad Fugius called BIM chaos.

00:05:08: Yeah,

00:05:08: inconsistent properties, different software defaults fighting each other.

00:05:11: It's a nightmare.

00:05:12: So what was the solution?

00:05:13: It sounded quite practical.

00:05:15: It was.

00:05:16: Instead of trying to fix it.

00:05:17: afterwards, he proposed creating a contractual IFC properties requirements table before anyone starts modeling.

00:05:24: Contractual.

00:05:25: So binding.

00:05:26: Yes.

00:05:27: Clearly defining what properties are needed, what format they should be in, and at which project stage.

00:05:32: It basically makes data standards part of the contract.

00:05:35: That's the kind of discipline needed.

00:05:37: That discipline connects directly to the broader idea of information control, doesn't

00:05:40: it?

00:05:40: It really does.

00:05:41: Albert Wu made the point that proper document control isn't just admin support, it's actually strategic.

00:05:47: Oh, so?

00:05:48: Because it ensures that single source of truth.

00:05:50: Central storage, automated version control, all that's crucial.

00:05:54: But importantly, it creates a transparent audit trail.

00:05:58: Who got what document?

00:05:59: and when.

00:06:00: Ah, the traceability.

00:06:01: Exactly.

00:06:02: Essential for compliance and honestly for resolving disputes later.

00:06:06: It voids a lot of, he said, she said.

00:06:08: Okay, thinking about disputes and ownership, Ralph Monague brought up the future, looking towards Web Three, that whole read-write-own idea.

00:06:15: What does that actually mean for construction data and liability?

00:06:18: It's potentially a game changer.

00:06:20: Current Web Two platforms, even good CDEs, are centralized.

00:06:24: They can still have issues with who truly owns the data, and liability can be murky.

00:06:30: Web three, with its decentralized ledgers, offers true data ownership and, critically, an immutable audit trail.

00:06:38: Every change, every distribution, is permanently logged and verifiable by everyone involved.

00:06:43: Immutable, so it can't be changed.

00:06:45: Correct.

00:06:46: That could fundamentally tackle the fragmentation and liability worries that current centralized systems struggle with.

00:06:52: It's a longer term vision, but very powerful.

00:06:54: Okay, so we've got cleaner data, hopefully, structured in a BIM methodology.

00:06:58: The next logical step seems to be adding time for D, connecting time and model data for that immediate visual payoff.

00:07:05: Definitely.

00:07:05: And David Millen stressed that forty visualization isn't nice to have anymore.

00:07:09: It's becoming a key requirement for successful delivery.

00:07:12: Why the shift to requirement?

00:07:13: Because it brings clarity instantly.

00:07:15: James Bowles broke it down nicely.

00:07:17: A forty model combines the schedule, the design, and logistics, seeing all three together visually.

00:07:21: Let's you spot clashes in time and space.

00:07:24: Precisely.

00:07:25: Teens can plan better, coordinate better, reduce bottlenecks, especially on complex sites.

00:07:31: It improves time management and cuts costs.

00:07:34: And the visuals themselves are getting seriously good.

00:07:36: Christa Jean Villabic highlighted the integration between Synchro-Forty Pro and Unreal Engine

00:07:42: V. Yeah, that's moving beyond basic block models into really high quality, almost photorealistic simulations of the construction sequence.

00:07:50: Which must help with communication, right?

00:07:52: Especially with stakeholders who aren't deep in the technical weeds.

00:07:55: Absolutely.

00:07:55: It makes it much more immersive, easier to understand.

00:07:58: And all this visualization helps front load certainty, particularly in pre-construction.

00:08:02: Finding problems earlier.

00:08:03: Way

00:08:03: earlier.

00:08:04: Sneha Kumari mentioned platforms like Merlin AI.

00:08:08: They centralize data, automate things like quantity takeoffs, apparently saving teams fifty-seventy percent of time there.

00:08:13: Wow, that's significant.

00:08:15: It

00:08:15: is.

00:08:15: And then they apply dynamic scheduling, risk analytics, to that four-D data.

00:08:19: The goal is to catch those fundamental constructability issues before you even break ground.

00:08:23: All right,

00:08:24: let's get to the site itself, where the digital meets the dirt.

00:08:27: On one side, you see amazing things.

00:08:29: Muhammad for Kambalaka talking about the precision, the speed, the safety of robots.

00:08:34: Yeah, some cool examples out there.

00:08:35: And Daniel Lorenzo showcased that incredible project, the first fully, three-D printed, compliant building done entirely on site with robotics and AI.

00:08:46: I mean, that sounds like the future has arrived.

00:08:49: It certainly looks like it sometimes.

00:08:50: But if

00:08:51: it's so great, why isn't it everywhere?

00:08:53: What are the real obstacles stopping this from scaling up massively?

00:08:57: Well, David Moser provided a really useful reality check, especially looking at solar construction robotics, but the lessons apply broadly.

00:09:04: Okay, what's the reality?

00:09:06: Site variability is a killer for robots.

00:09:09: and even ground, different soil types, weather changes.

00:09:12: Robots struggle with inconsistency.

00:09:14: You like predictable environments.

00:09:16: Exactly.

00:09:16: Then there's this huge challenge of integration, getting multiple specialized robots to actually work together smoothly.

00:09:22: And the mobilization costs are often really high.

00:09:25: So the efficiency gains might not be as huge as promised initially.

00:09:29: Often, yes.

00:09:31: The timelines can be overly optimistic because these factors aren't always fully accounted for.

00:09:36: That probably explains the lag errantry living observed in the UK.

00:09:40: He said AI use in the office is like, fifty-two percent, but only fifteen percent on actual sites.

00:09:45: Big gap.

00:09:46: Yeah, and the excuses are always cost or data readiness, but maybe it's just that fundamental difficulty of translating digital plans into messy site reality.

00:09:55: It could be.

00:09:56: And this is where we absolutely have to circle back to basics.

00:09:59: Erdem Everyn made such a critical point.

00:10:02: Technology is Basically useless if you ignore project management fundamentals.

00:10:06: You mean like scheduling communication?

00:10:08: Yes,

00:10:09: but even more basic.

00:10:10: He listed four breakdowns that quietly kill projects no matter what tech you use.

00:10:14: Accepting delays as normal is one.

00:10:16: Wait, accepting delays as normal doesn't everyone do that?

00:10:19: Sadly, yeah, it's common.

00:10:21: But Evern argues it's a failure of discipline.

00:10:24: Other killers.

00:10:25: Misusing schedule flow.

00:10:26: Approving submittals after the work is installed.

00:10:28: Okay, that's bad.

00:10:29: Right.

00:10:30: and skipping proper crew briefings.

00:10:32: His point is, discipline controls.

00:10:35: Managing the basics rigorously, that's what really drives efficient delivery.

00:10:40: No fancy tool.

00:10:42: fixes fundamental process failures.

00:10:44: That's a really grounding perspective.

00:10:45: Tech isn't a magic wand.

00:10:47: Okay, finally, let's link this back to sustainability.

00:10:50: Hubert Romberg had an interesting take.

00:10:52: Yeah, he said transformation comes from scaling knowledge, not just scaling individual companies.

00:10:57: Scaling knowledge, how?

00:10:58: Well, his company, Cree Buildings, focuses on open sourcing their proven sustainable methods like their timber hybrid system.

00:11:06: They share the knowledge.

00:11:07: No

00:11:07: others can use

00:11:08: it.

00:11:08: Yeah, exactly.

00:11:09: And their partners, using these shared methods, reportedly saw forty-three percent faster execution and lower operational costs.

00:11:15: It's about accelerating change by sharing what works, not hoarding it.

00:11:19: A collaborative approach to scaling innovation.

00:11:21: Interesting.

00:11:22: Hashtag, hashtag, outro.

00:11:24: So if you step back and look at all these insights from weeks thirty-eight and thirty-nine, the big message seems pretty clear.

00:11:29: The company's really succeeding, aren't just grabbing shiny new tech tools.

00:11:33: It's more integrated than that.

00:11:34: Yeah, they're focused on convergence, how AI connects to clean data, how that fuels better BIM standards, how that drives effective four D planning.

00:11:42: And critically, they're applying fundamental discipline to how they manage projects.

00:11:48: Rubensor has summed it up well.

00:11:50: Prioritize clear outcomes, real value, not just the latest features.

00:11:54: That focus on fundamentals feels key, especially when you hear things like Jihao Zhao noting that Excel is still the dominant design tool in AEC in twenty twenty five.

00:12:04: Still Excel.

00:12:05: Right.

00:12:05: While Victor Machiri points out the huge opportunity in just vectorizing old two D drawings into usable three D models.

00:12:12: So maybe the final thought for you, the listener is this, how are you pushing to finally bridge that gap between the two D contract world we often live in and the three D model dream we talk about?

00:12:21: Good

00:12:22: question to Ponder.

00:12:23: If you enjoyed this.

00:12:24: dive new segments drop every two weeks.

00:12:26: Also check out our other editions on smart manufacturing and digital power tools.

00:12:30: Thank you for diving deep with us and we encourage you to subscribe.