Best of LinkedIn: Digital Construction CW 42/ 43

00:00:00: This deep dive is provided by Thomas Allgeier and Frennus based on the most relevant LinkedIn posts about digital construction in calendar weeks forty two and forty three.

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.

00:00:22: Welcome to the deep dive.

00:00:24: Our mission today, well, it's really to cut through the noise, isn't it?

00:00:27: We're looking at the top digital construction trends, the key lessons, and maybe some impactful insights shared over the last couple of weeks.

00:00:34: Yeah, exactly.

00:00:35: And looking across all the posts, there's definitely a vibe, a theme emerging.

00:00:39: Pragmatism?

00:00:40: Languageism,

00:00:40: okay.

00:00:40: Yeah, it feels like a real shift away from just chasing, you know, the next shiny thing, the hype.

00:00:45: The focus now seems much more on, let's say, targeted execution.

00:00:49: So

00:00:49: making it work in the real world.

00:00:50: Right.

00:00:51: How do we pair these digital tools with actual workflows, day-to-day stuff?

00:00:56: And crucially, how do we get measurable results?

00:00:59: Like that question Ayub El Amrani cited, how does this save me two hours at eleven

00:01:04: p.m.?

00:01:04: That really hits home.

00:01:05: It does.

00:01:06: It's moving from theory to utility.

00:01:08: So, let's dive into where that strategic focus is landing first.

00:01:13: And the big one, obviously, is AI.

00:01:15: Right.

00:01:16: AI.

00:01:16: Everyone's talking about it.

00:01:18: For sure.

00:01:19: The excitement's palpable.

00:01:20: But interestingly, the consensus from a lot of the key voices we saw is, well, strategic caution.

00:01:26: Caution.

00:01:27: Not just jump in.

00:01:28: Sort of.

00:01:29: The main advice seems to be, let's stop looking for that one magic bullet.

00:01:33: Okay.

00:01:34: Like Daniel Spink was saying, don't expect some perfect AI system to just run everything for you.

00:01:39: Exactly that.

00:01:40: His suggestion was, let's learn from the early BIM days maybe.

00:01:43: Start slow.

00:01:44: Right.

00:01:44: Pick specific areas like finance maybe.

00:01:47: Automate some simple data collection first.

00:01:49: You know, build that trust within the organization before you try to scale it everywhere.

00:01:52: That makes sense.

00:01:53: And I guess the reason for that caution ties back to the industry's, well, perpetual headache.

00:01:58: Data.

00:01:58: Ah,

00:01:59: the data barrier.

00:02:00: Yes.

00:02:00: We saw AJ Waters, Mittle Sharon, Elliot Christensen.

00:02:04: They're all really hammering this point.

00:02:05: AI is only as smart as the fuel you give it.

00:02:08: Simple as that.

00:02:10: Mittle Sharon put it quite bluntly.

00:02:11: If your process is broken on paper, technology just breaks it faster.

00:02:15: Ouch.

00:02:16: Yeah, the foundation issue.

00:02:18: If your current way of working is chaotic, the AI just learns chaos.

00:02:22: Precisely.

00:02:24: So suddenly, clean, structured, connected data isn't just nice to have, it's like the core competitive advantage.

00:02:30: Okay, so let's say you do have clean data.

00:02:33: What then?

00:02:33: What makes AI actually work?

00:02:35: Right, that's where context comes in.

00:02:37: Doug Vincent laid out a really neat recipe for it.

00:02:39: Oh yeah?

00:02:39: He said it's AI plus context plus data.

00:02:42: AI plus context plus data.

00:02:44: Yeah,

00:02:44: AI on its own.

00:02:45: It's a bit surface level.

00:02:47: But when you layer in that context, the AI, knowing who you are, what your job is, the specifics of the project.

00:02:53: Then

00:02:53: it gets useful.

00:02:54: Then it transforms.

00:02:55: It becomes almost like a hyper-specific agent, a collaborator tailored right to the job site or the office task.

00:03:01: OK, but hang on.

00:03:02: If context is so critical, does that mean we lose the power of those big, general AI models?

00:03:07: Are we just reinventing custom software?

00:03:09: That's a really good question.

00:03:13: But maybe the skill shift that Omri Stern talked about addresses that.

00:03:17: What was that?

00:03:18: He reckoned the next wave of skilled workers won't necessarily be coders.

00:03:21: They'll be prompt engineers.

00:03:22: Clomped

00:03:23: engineers.

00:03:23: Yeah,

00:03:24: project pros who can actually translate the logic of the field, the complexities of construction into machine logic.

00:03:32: Basically, people who can guide these AI agents effectively to give them the right context.

00:03:37: Interesting.

00:03:37: So bridging the gap between the site and the system.

00:03:40: Exactly.

00:03:41: And you know, what really stood out for me in all this AI talk wasn't just the future potential, but where Jackson Rose said the impact is actually landing right now.

00:03:49: Oh,

00:03:50: where's that?

00:03:51: Not robots on site yet.

00:03:52: Not primarily no.

00:03:53: He said, the immediate value of the real actions is happening in the owner's office.

00:03:57: The owner's office?

00:03:59: How so?

00:04:00: Streamlining admin tasks, checking contract compliance automatically, improving reporting, making knowledge easier to find.

00:04:07: OK, so freeing up the owner's team for more strategic stuff.

00:04:10: Precisely.

00:04:11: That's a massive, tangible ROI for capital programs happening today, not tomorrow.

00:04:16: Right.

00:04:17: OK, so if clean data is the fuel for AI, then things like BIM and digital twins, they're kind of like the engines using that fuel.

00:04:26: That's a great way to put it.

00:04:27: And yeah, these tools are definitely moving beyond just being static, three D models.

00:04:30: They're becoming predictive.

00:04:32: But again, that data foundation is absolutely everything.

00:04:35: It always comes back to the data.

00:04:37: Always.

00:04:38: Nicola Jovic really emphasized this for BIM.

00:04:40: He said projects often fail early because teams get excited and chase quick wins instead of being disciplined at the start.

00:04:47: So rush the setup.

00:04:48: Yeah,

00:04:48: he basically said the goal initially shouldn't be victory.

00:04:51: It should be clarity.

00:04:53: Clarity.

00:04:53: What does that mean in practice?

00:04:55: Alignment.

00:04:56: Getting standards agreed upon, process is clear across the whole team.

00:04:59: You know, build the foundation before you build a house.

00:05:01: Clarity over speed.

00:05:03: Sounds simple, but probably hard-earned wisdom.

00:05:05: Okay, what about digital twins then?

00:05:07: DTs.

00:05:08: Right,

00:05:08: DTs.

00:05:09: Santosh Kumar Boda shared some compelling numbers.

00:05:12: He confirmed they deliver hard ROI.

00:05:15: How much?

00:05:15: Like, ninety-two percent of companies using them report over ten percent ROI.

00:05:20: Often specifically by predicting failures before they happen.

00:05:22: Wow, okay, that's significant.

00:05:25: predicting failures.

00:05:25: Real money saved.

00:05:27: But the technical hurdle is still pretty huge.

00:05:30: Fluorine humor pointed to the number one killer of DT projects.

00:05:34: Which is?

00:05:34: Data pre-processing.

00:05:35: Ah, cleaning the data again.

00:05:37: Exactly.

00:05:38: Raw sensor data.

00:05:40: It's messy.

00:05:40: It's got gaps, inconsistencies.

00:05:42: It needs constant cleaning, normalizing.

00:05:45: That prep work, he called it the crucial fuel for the DT engine.

00:05:49: And getting it wrong, that's apparently what scuppers most projects.

00:05:51: So

00:05:52: the unglamorous work is actually the most critical.

00:05:54: The story of construction, maybe.

00:05:56: Could be.

00:05:57: But on a slightly more accessible tech note, Christian John Velivik shared something cool about visualization.

00:06:03: Yeah.

00:06:03: They used game engines for a four-D visual planning.

00:06:06: Game engines, like for video games.

00:06:08: Yeah.

00:06:08: And the effect was it made these complex schedules instantly understandable, even for people in the team who'd never read a Gaunt chart before.

00:06:15: They could suddenly see the plan, spot potential clashes, and actually collaborate on

00:06:20: it.

00:06:20: That's powerful.

00:06:21: Making complex info intuitive.

00:06:23: It really is.

00:06:24: Shows how tech can solve basic communication issues.

00:06:27: But just to circle back to those stubborn challenges, Fabio Bronson reminded everyone that fully automating scan to BIM.

00:06:36: Still a problem.

00:06:37: Still one of the biggest unsolved costly manual headaches in contact, apparently.

00:06:42: That nut is proving really tough to crack.

00:06:46: Okay, so these challenges, like Stand to BIM, they obviously create opportunities for the tech market, right?

00:06:51: Startups, big players, they must be responding.

00:06:55: Absolutely.

00:06:56: Which kind of brings us to the tools and platforms themselves.

00:06:58: How are we actually getting this stuff done in the field?

00:07:01: Let's talk about that.

00:07:01: What's happening in the ecosystem?

00:07:03: Well,

00:07:03: there was some significant market movement recently.

00:07:06: Arvin Veluvali pointed out Procore's launch of Helix, their new AI agents, and for developer portal.

00:07:13: That sounds big.

00:07:14: He called it basically dropping an atomic bomb on specialized tech startups.

00:07:18: Wow, why?

00:07:19: Because it signals this major shift, right?

00:07:21: Integrated platforms are now building in native AI and opening up their systems that could really disrupt those single point solutions almost overnight.

00:07:28: It's interesting times for the smaller players then, but they're innovating too, surely.

00:07:33: Yeah, for

00:07:33: sure.

00:07:34: Alistair Lewis actually proposed a kind of counter move, an alternative AI tech stack, specifically for smaller architecture practices.

00:07:42: aimed at cutting costs.

00:07:44: Exactly.

00:07:45: Cutting costs and avoiding getting locked into legacy tools.

00:07:48: He suggested a low-cost cloud-based stack using things like FVO-DD for scanning, draft for modeling, speckle as a data hub.

00:07:57: So, accessible collaborative alternatives are popping up.

00:08:00: Yeah,

00:08:00: there's clearly a demand for that.

00:08:01: And then out in the field itself, we're seeing automation where it makes obvious sense.

00:08:06: Ivo Van Broekland highlighted robotic layout systems.

00:08:08: Like robots drawing lines on the floor?

00:08:10: Sort of, yeah.

00:08:11: But integrating directly with the BIM data to do super precise markings on site automatically.

00:08:17: Okay, I can see the benefit there.

00:08:18: Accuracy, speed,

00:08:20: less rework.

00:08:20: Big time.

00:08:21: Cuts down rework, speeds things up.

00:08:23: And thinking longer term about automation Patrick Hellerman had an interesting take on industrial humanoids.

00:08:29: Those general purpose robots we hear about.

00:08:31: Yeah, the expectation might be they'll do everything.

00:08:35: But he speculated they're more likely to evolve into specialized tool

00:08:39: carriers.

00:08:40: Automated platforms for the tools humans already use.

00:08:44: Maybe the reality of specialized construction work means the universal robot is still a way off.

00:08:50: interesting perspective.

00:08:51: Okay, so we've covered the tech, the data, the tools, but none of that works without the right processes and well, people.

00:09:00: Absolutely.

00:09:01: Which leads us neatly into our final theme.

00:09:03: Lean strategies, collaboration, and that crucial element culture.

00:09:07: How do you actually maximize the investment in all this digital stuff?

00:09:11: Right.

00:09:11: It's where the change actually sticks, or doesn't.

00:09:14: Perry Thompson shared a brilliant, really down-to-earth story about this.

00:09:17: Oh, yeah.

00:09:17: It was about an electrical crew at Parsons Electric.

00:09:20: They implemented the Five S system on site.

00:09:23: Five S?

00:09:23: Sort.

00:09:24: Set in order.

00:09:25: Shine.

00:09:25: Standardize.

00:09:26: Sustain.

00:09:27: Yeah.

00:09:28: All about getting organized.

00:09:29: And apparently the crew complained at first, you know, this is just extra work.

00:09:33: But Thompson said by week three they were actually ahead of schedule.

00:09:36: Ahead?

00:09:37: Just from being tidy.

00:09:38: Well

00:09:38: think about it.

00:09:39: They stopped losing tools, stopped walking back and forth to the truck for things they forgot.

00:09:44: Stopped waiting for materials because things weren't where they should be.

00:09:49: So they spent more time actually working.

00:09:51: Exactly.

00:09:52: Less time searching, more time doing.

00:09:54: It's such a simple example.

00:09:56: But it shows lean isn't just for factories.

00:09:59: It pays off right there on site.

00:10:01: That's a great anecdote.

00:10:02: Simple discipline makes a difference.

00:10:04: And speaking of discipline, improving operations also means changing how decisions get made.

00:10:10: Right.

00:10:10: Moving away from just gut feeling.

00:10:13: Definitely.

00:10:13: Felipe engineer Manarricas was championing a method called choosing by advantages or CBA.

00:10:19: CBA.

00:10:20: How does that work?

00:10:21: It's a systematic way to compare options.

00:10:23: The really interesting part he noted is that it focuses on the actual beneficial differences, the value between alternatives first, cost.

00:10:30: That's kept separate until the very last step.

00:10:32: Ah, so you don't automatically default to the cheapest option without thinking about the real value?

00:10:37: Exactly.

00:10:37: It forces a focus on value before price.

00:10:40: Smart.

00:10:41: And that kind of systematic thinking, I guess it ties into the bigger picture of collaborative contracts.

00:10:45: Absolutely.

00:10:46: Donanda made a strong point about that.

00:10:48: Things like progressive design build or IPD.

00:10:51: They only really work, he argued, if they require certain elements.

00:10:56: Like what?

00:10:56: Like co-location, having teams physically together, using target value delivery, TVD.

00:11:02: and actually deploying lean, BIM, VDC principles as part of the contract.

00:11:08: It's gotta be an integrated package deal.

00:11:10: Not just a different contract template.

00:11:11: Right.

00:11:12: And finally, thinking about making all this change happen, especially for the younger professionals trying to push things forward in AEC, Ricardo Kahn offered a framework.

00:11:21: Based on Cotter's steps for change.

00:11:23: Yeah,

00:11:23: Cotter's eight steps.

00:11:25: But the key takeaway he highlighted was focusing on generating small, visible wins early on.

00:11:29: Short-term wins.

00:11:30: Exactly.

00:11:31: Find something you can improve, quantify it, hour saved, errors avoided, and then shout about it.

00:11:36: Share that success.

00:11:37: Build

00:11:37: momentum.

00:11:38: Build momentum.

00:11:39: Build acceptance.

00:11:40: That visibility is what helps shift the culture over time.

00:11:43: Makes sense.

00:11:44: Technology might be the vehicle, but it feels like people on process are definitely the engine driving it.

00:11:49: Well said.

00:11:50: If you enjoyed this deep dive, new editions drop every two weeks.

00:11:54: Also check out our other editions on smart manufacturing and digital power tools.

00:11:58: And maybe something to think about as you process all this.

00:12:01: We heard so much about clean data, about collaboration, about the need for trust, especially for AI to really work.

00:12:08: So here's a thought.

00:12:09: If AI success really hinges on trust and sharing clean data.

00:12:14: How do our industry contracts and data sharing agreements need to fundamentally change to actually encourage that collaboration?

00:12:22: That's a big question.

00:12:23: Definitely something for you to mull over until our next deep dive.

00:12:27: Thanks for listening and don't forget to subscribe for more insights.