Become a Creator today!Start creating today - Share your story with the world!
Start for free
00:00:00
00:00:01
E4: Trains too big for the station?
68 Plays3 years ago
In this episode Carlos & Jason chat about the Spanish Transport Secretary's blunder earlier in the year, and how issues are unavoidable when systems used in construction don't 'talk to each other'.
They also debate whether software or hardware is more challenging to develop in construction, and what the big players like Autodesk & Procore are doing right.
Recommended
Transcript
Introduction to the Offsite Podcast
00:00:00
Speaker
Hey everyone and welcome to the offsite podcast where we chat all things construction, technology and everything in between. My name's Carlos and I spend most of my days talking to construction teams about how they deliver projects. And I'm Jason and I build software for construction teams.
00:00:16
Speaker
So today
The Spanish Train Size Fiasco
00:00:17
Speaker
we're going to be talking about the resignation of the Spanish Transport Secretary, the acquisition of a company called Easy Power by Bentley, and some safety technology being tested by Ferrovial. So first up, the Spanish Transport Secretary. Lots of news on this at the moment.
00:00:33
Speaker
mainstream news is actually hitting things like the BBC, not just our construction channels that we follow. And I guess in a nutshell, the Spanish transport secretary, who's also the head of the sort of rail company or the state run rail company, has resigned after they've realised that they've spent 200 million euros on dozens of trains, which actually won't fit in a bunch of tunnels in, I think, the northern region of the country.
00:01:02
Speaker
unfortunate
Construction Specification Challenges
00:01:03
Speaker
day in the office. Yeah that is a stressful day in the office. Obviously a complete nightmare, probably quite easily done. I'd imagine first off they probably don't have great asset information and like obviously they're not going to be nowhere near things like digital twins for infrastructure that's potentially hundreds of years old.
00:01:25
Speaker
So you can imagine there's this big old problem of trying to understand what's out there and get that information in one place to be able to effectively design, but that is a big old fuck up.
00:01:36
Speaker
I, I thought for sure you were going to go in and, and, um, and like attack the, this poor chap and the, and the, um, and the, and the Spanish. Um, but as someone that's someone that's built a bridge deck 20 mil too wide, uh, one time on my initial thought was like, what a poor fucker. The, uh, the, my.
00:02:01
Speaker
I guess I kind of want to intuitively defend them because without knowing anything about the situation, I do empathize with the fact that I would imagine that the way that in construction specifications are done can be a complete mess.
00:02:21
Speaker
I guess it stems from, you know, you get a situation where you'll have like a performance spec that says X and a prescriptive spec that says Y and they're buried across five different documents and they're kind of the way we do specifications and construction is just like a, is a, is a mess. And oftentimes it stems from like the client, I guess, not fully knowing what they want because the design's not developed so much that they can be prescriptive about it.
00:02:48
Speaker
I can think of so many previous projects where in order to order a window, you needed to dig out requirements from four different documents. One saying one thing, one saying the complete opposite, and then one having a performance specification that was totally incompatible with the two things that were in the other places.
00:03:08
Speaker
Do you think you have empathy though because you've been through something like that because from an outsider perspective it's hundreds of millions of pounds worth of taxpayers money and it's a trade not fitting in a tunnel. It's not like a little design tolerance.
00:03:22
Speaker
I am making the wild assumption that there isn't somewhere in the specifications that says train shall not be bigger than X because if it does say that, then I take back everything that I'm saying. But my
Ensuring Train and Tunnel Compatibility
00:03:38
Speaker
guess is that there's like a combination of, you know, the train shall be big enough to fit in all the tunnels, which tunnels, which ones are we, you know, like,
00:03:47
Speaker
I can imagine it's a very complicated, like, an easy statement that's kind of very, very difficult to validate. Yeah, maybe not. I can imagine, like, I'd imagine in London, if you were building a bridge and you told TfL it's x meters high, they'll go, nah, no buses fitting under that. Surely there's someone somewhere that, like, yeah, someone screwed up. They've obviously, someone wanted a head and the head of the
00:04:13
Speaker
state rail company has been sacked probably had nothing to do with this but
00:04:18
Speaker
The opposite is easier though. Like when you're building the tunnel, it's kind of easier to go. What are the set? Yeah. What's the biggest train or what's the biggest, you know, bus that needs to fit under it. It's actually like kind of hard to do the opposite when you're buying it, especially if you don't have like, again, TFL is super organized and on it, but like you could easily imagine if you or I ran the transport network there.
00:04:46
Speaker
It would be a different story. It'd be like, send someone to measure all the tunnels, please, to work out what size we should order. Yeah. Just take your biggest train and run it across all tracks and see if it makes it. What you should do is get your smallest train and then coat it in some sort of, like, material and then just drive it through all the tunnels. Yeah, drive it through all the tunnels. Yeah, like a cut-out. Find your smallest one and then that's it.
00:05:16
Speaker
That's what you should do. It's good that we're talking about safety later because that's a good solid paper. Next time you go to Spain and you just see this giant white train doing laps around. Oh, that's the new polystyrene train that we're doing. So that we should have, we should paint about it. Polystyrene is even too hard. It'd have to be even softer. So like, I don't know, like kind of like a big marshmallow or something.
00:05:42
Speaker
It would be like those, you know, those kid toys where you like put your hand on and it pushes the pins out. Oh, that one where you get those videos on TikTok where they're like chopping it and it's really soothing. It'd be that, whatever that is. It's some kind of like sticky sand. Chloe will know what it is. She'll look up and put it in the notes. Yeah, I'm not a TikTok guy. No, neither am I. In terms of modern tech now,
00:06:10
Speaker
If we were 100 years into the future, would a digital twin resolve this because you've got this digital asset? It would probably take 100 years until the end. You didn't get five years into the future and 100, okay. No, no, yeah. But I mean, you can imagine we'll get to the point where there'll be government-owned digital assets of entire countries or cities. That should stop things like that, right? Because it's like a giant bin model and it would just say, this doesn't work because it's gonna clash here, right? That's where we're trying to get to.
00:06:39
Speaker
100% valid use case. Like that would, you would imagine that would completely solve the problem. Yes. Good solution. That or the giant marshmallow thing, you know, just yet. Yeah. Yeah. We actually had a similar issue on crossrail. I won't name the station, but the station I was working on at the time. I think you've said we have this issue where every episode prior, but yeah. If they're listening to that video, it's fine.
00:07:07
Speaker
the we had an issue where the alignment of the track against the platforms was off and the track was sort of laid by the design system that was managing the tunnels but the platforms were the design system of the station and there was a very slight sort of
00:07:25
Speaker
um skewed which meant that the doors were going to clash with the platform and the train was sort of ended up having to cut back the platforms in certain areas to make sure it could actually fit um i don't know if that's a system wide thing but um i guess it's the opposite sort of problem which the trains are fine but the actual system didn't work so um yeah these things obviously do happen quite a lot
00:07:48
Speaker
Yeah, on a project I was thinking of before, we were trying to purchase windows and there was a very prescriptive makeup of the windows. They were sort of 32 mil thick, five sheets of glazing with layers in between.
00:08:05
Speaker
And they had a very prescriptive specification and then a performance specification on thermal performance and blast performance. And we spent months going in circles because you couldn't meet all three of these things.
00:08:20
Speaker
And there was no budge. Well, we want all three things. So I guess I'm imagining, I'm hoping for the sake of the, the Spanish gentlemen, that it's a, it's more of a situation like that. But I guess the, um, the news doesn't care. Yeah. The news is news. It's, uh, yeah. You can just change the name.
00:08:41
Speaker
Yeah. Right.
Bentley Acquires Easy Power
00:08:44
Speaker
So next up, Bentley have acquired a company called Easy Power. Easy Power is a tool that allows you to produce like electrical schematic type drawings. I've got some experience in that from a previous M&E based role. And these drawings, they're like their proper old school. They look like a flow diagram of different systems and how they connect with each other and what that sort of is laid out like.
00:09:11
Speaker
The company is a 40 year old company, which I don't think anyone's ever heard of. So the headlines Bentley acquires middle aged man. Yeah. So I don't know. I read it and I thought, are they like, is their business model now so lazy that they go? Oh, there's this tiny gap in our portfolio somewhere. So let's just buy a company rather than try and do something that's better, build their own, do something about it.
00:09:37
Speaker
I'm pretty sure Auto can have their own version, so maybe they try to compete with them. But it seems like an odd one for me.
00:09:45
Speaker
But I guess I'm not familiar enough with the electrical schematic during market. I don't know what's going on here. These, I guess, older or legacy, not legacy, that's too aggressive, but like older construction tech companies that purchase a million different, they have a million different brands. Specifically, we happened to reference Bentley last week.
00:10:09
Speaker
I, yeah, it just leads to a completely disconnected fragmented set of tools that don't talk to each other. And it's like, I think it's everything that's wrong with construction software and technology. But what I don't understand, I can't understand the business logic in purchasing like a piece of software like that. You know, if you are a software company or a technology provider, you know, when you are building products, it takes
00:10:33
Speaker
It takes so much effort to build the first version of something to invent the solution to the problem and to find the solution that fits and actually solves the market problem. But once someone's done that, it's so easy to copy it. So I don't understand why if you are a company that does have a core competency in building software, you don't just build a solution.
00:10:57
Speaker
that is integrated with your tools. Unless your core competency isn't building software, it's just buying them and selling them. Yeah. You touched on there about connecting these other apps. There was an article of Bentley that says the next step is to connect you with these eight different platforms or something like that, which is crazy.
00:11:20
Speaker
Then the, I guess, Autodesk Oracle, they do typically purchase tools rather than build. Do you think part of it is they use resellers? They're not, they haven't got their ear to the floor in terms of customers. They're just sort of this massive organization that's just dishing out licenses for tools. Do you think that they actually would struggle to build a tool? Because maybe they're not full of sort of construction experts, they're full of sales people,
00:11:46
Speaker
commercial things like that. I don't know because some of their tools they do invest time in like their Synchro product they invest time in they do like you see case studies all the time where they're making adjustments and improvements based on feedback of customers. We saw a case study recently from I want to say it was Hinckley or another project. There was another one from recently from Melbourne from the level crossing removal project down there.
00:12:14
Speaker
So they do, but I, you know, it's hard when you've got, I can't remember what the number was. It was like over a hundred products or something that we looked on their website. So some of them are just sitting there on the shelf for the people that still purchase them. And maybe they have a set of like,
00:12:32
Speaker
Here are the ones that are the future and we're investing in those. And there's a bunch of other ones that just sit on the shelf, but they need to be investing it because if you, if you look at the space, like Autodesk and Procore are not sitting still. They're moving very, very quickly to build a properly connected platform and you know, 120 tools or whatever in different isolated spaces is just not going to cut it. Yeah.
00:13:01
Speaker
The really crazy one was Aconex being purchased by Oracle, was it 1.2 billion or something like that? That was a long time ago as well. Ten years ago now, something like that? Must be. It's not like a crazy complicated product, so to spend that much money on something
00:13:22
Speaker
Yeah, but they said users, they were acquiring the users and into the ecosystem. But it's still kind of not connected to their other products very well. You take the other approach where Autodesk purchase plan grid must be, I think it was 2018 or so. So that's five odd years ago. And I think they spent 800 million on that acquisition.
00:13:44
Speaker
and it took some time and playing grid was standalone and I think they still have standalone playing grid customers but it's largely being integrated into the new construction cloud now in kind of a nice way from what I've played around with.
00:14:00
Speaker
And like, you know, that is a way easier, I think a way better proposition if I was a construction company than a collection of, you know, yeah, other platforms, other brands that aren't named, whereas just like a collection of apps that don't talk to each other.
00:14:18
Speaker
Yeah. Oh, the names on the battery website with nuts is like there was like 15 apps and all of them started with the word open. Don't really recognize this opens good. It's like I stick the letter I in front of something like the IV or the open viewer. Both of those are positive.
00:14:36
Speaker
Yeah, better off sticking Mac on the end of each word. I will remember then. Right. Last up, Ferrobial.
New Safety Technologies in Construction
00:14:43
Speaker
So they are testing a new safety system. It's a proximity system where you stick sort of sensors or little devices on your vests, helmets, plan, all sorts of things that are moving around site. And ultimately, if two of them come too close together, the sort of machine gets shut down and a siren goes off and there's like a light that sort of
00:15:06
Speaker
a red light that's very visible. There's a lot of sort of tech in this space. So it's interesting to see Ferrovial. I was actually a subsidiary of Ferrovial that sort of built this and deploying this tool and doing it themselves. For me, the tech seems quite simple. So it probably does make sense for a big company to just build it and install it and cost their sites rather than maybe pay a supplier or a third party a lot of money continuously.
00:15:35
Speaker
each year. I don't know. Hardware is notoriously difficult. So I would, my initial reaction is like, it's a very complicated thing to build. I'm very surprised that they're building something themselves. Hardware is like... Why is it that, I guess, I'm not a specialist in any way of building hardware. But it's like, it's literally just a proximity sensor. And when two things get too close together, it pops an alarm off and flashes a light.
00:16:05
Speaker
do you know how you know how we build software Carlos and occasionally very very occasionally something goes wrong and when it goes wrong we very very magically and quickly fix it we can do that because it's software hardware is
00:16:21
Speaker
hardware has like a whole bunch of problems like yeah there's a whole amount of testing that needs to be done you have to also be thinking well into the future because if you sell or purchase a lot of sensors and then you think of another feature and these aren't compatible with this new feature you've got a bunch of like stranded hardware assets there's a whole bunch of additional safety considerations with a piece of hardware than it is with software reliability testing there's a hardware is like
00:16:47
Speaker
substantially harder to do. And in this example, the consequence is high if it's wrong. So yeah, exactly. Yeah, absolutely. You know, who knows the space really well, we should get we should talk to Tommy from links. Obviously, they have a similar type thing. I think they provide like the network
00:17:12
Speaker
as well as the sensors because you need a network for these things to run in to feed information. Yeah, you need like 5G systems across the site and things like that, right? Yeah, 5G. Well, then I don't know what the... Yeah, there's a whole thing of like battery life and then how to just charging work. Hardware is well hard, is the summary. Agreed.
00:17:33
Speaker
I guess what's interesting, because you're talking to teams like every day. Yeah, like you said, it is kind of, there are a few of these providers that you see around the place. Do you see much of it in like use on a project when you're talking to teams?
00:17:53
Speaker
I haven't seen one particular vendor or service used widespread, but I come across a lot. There's obviously Plinks that we mentioned. There's a company called SiteZone. There is a company that's actually owned by JCB or one of the big plant and I companies, and they have their own systems.
00:18:12
Speaker
every site seems to have the old school reversing sort of senses and noises for plant, not necessarily proximity to people, the people seems to be this sort of neurodiction. And it's a hard one for a site to say no, it's kind of like, who do you go for, not do you use it, because it's safety, it's quite obviously important. So I'd imagine we're going to see
00:18:39
Speaker
one of these guys really sort of dominate the space. But at the moment I just keep seeing different companies. And obviously I've just mentioned three, four OBLs a fourth now. So there's no sort of common. Do you, do you like actually see it? Like I just not to say, not to, I guess, say it will or won't happen, but I guess stay to play at the moment. I, because I, I don't know. I haven't seen it myself, but do you see it in use on projects like today? Or do you think it's like something that's coming?
00:19:10
Speaker
It's definitely on a bunch of sites that we work with and I've been to now, so it is being used. I don't think it's trickled down to your five, 10 million pound projects, but your big public multi-billion pound projects for sure are using this.
00:19:30
Speaker
that when you say using it is like, I imagine there's like levels to the complexity. So there's like interactions between people and plant. And you could imagine quite a dumb device quite easily could do that just based on proximity sensing between the two. Yeah. The sort of dumb ones are if someone walks near a piece of plant, it goes off.
00:19:52
Speaker
Yeah, there's nothing on the human. It's just on the plant and it's got like a sensor that can see if something's within two meters or whatever that is. We actually had that back on crossroad. Yeah, I remember that from yeah, exactly. I remember that from ages ago. Because I know that the one that this couple that I've seen, you know, the people are wearing them, they're somewhere in the helmet.
00:20:14
Speaker
Um, yeah. And they're tracking the kind of a combination of there's a, there's a safety aspect, like proximity sensor to it, but there's also like, um, I don't know if you have this in football, but over, over in Australia, like, uh, if you watch a rugby league or rugby in your match, they've always got like a sensor on there, uh, in the, in their jersey at the back, that's tracking how many meters they run, how many titles they make.
00:20:40
Speaker
Yeah, like on the pitch, so there'll be like how many tackles and how many like things sensing like and getting their like game stats. And I've seen some that are trying to do that on people. So there's like a safety element, like a productivity component to it. Yeah. BAM are trying to, they're sort of working in this space. They're testing something called BAM cam. So on a few of their sites, they've got kind of,
00:21:08
Speaker
I've got these cameras on a 5G network. You don't want on this spell that searching in Google.
00:21:22
Speaker
They've got a 5G network set up on a couple of sites and they've got these cameras that are basically picking up people and plant. So it's sort of tagging them and monitoring them. So they want to get to a point where they can confirm this bit of excavation. We had two pieces of plant and six people working for X hours and things like that.
00:21:41
Speaker
But I think they're trying to sort of build in safety to that too, so they can sort of monitor that and maybe have like a control room or control hub that can see sort of everything moving in real time.
00:21:52
Speaker
That's got to be difficult with a site that's constantly changing line of sight. Someone parked on a machine in the wrong spot and all of a sudden you can't see half the site. Just the volume of cameras. Imagine a high speed to a high wage job.
00:22:14
Speaker
Yeah, yeah. Just like you're in like a 10 metre channel for four kilometres. Like, what do you do? We'll see how they go in that space. I'll see if they want to rename it. That's all we've got time for today. So as always, thank you very much for listening.
00:22:37
Speaker
The greatest technology. Yeah. Thank you very much. I should have just said thank you very much. I won't say it.