The Role of Data in Construction with Josh Symonds

Andrea Ciaffi

Hi, Josh, how are you? Welcome to the Digital Construction podcast.

Josh Symonds

OK, alright, hang on. Could be here.

Andrea Ciaffi

It it's great to have Josh Salmon. Today is A is a great data leader. I worked with him in the past, that Arab and at the moment he runs his own business. We'll probably mention that during this interview and today we talk about Dada.

Andrea Ciaffi

So Dada is a, it's a big topic for the digital transformation and I'm very excited to have you, Josh today to talk about that with me. But before we start, I have a couple of eyes Breakers for you.

Josh Symonds

OK.

Andrea Ciaffi

And so again, relax you a bit and. Get started. So if you have to explain, you know like data could be a complex things and people don't understand what's going on. So I I thought a good icebreaker would be about asking you if you could display in your job to a five years or five years old.

Andrea Ciaffi

How would you describe it?

Josh Symonds

It's it's a pretty tough one to be, to be completely honest and always struggle. I've got a nine year old and a three-year old, so trying to explain to them what dad does. He's always been tricky. I think they they still have that perception that I just do like IT support.

Josh Symonds

That's got a lot of computers and that kind of stuff. So I usually revert back to my, to my old GIS role where I kind of just go. I do Google Maps on steroids or I do like advanced Google Maps. And that seems to get people across the line.

Josh Symonds

So I think if I was to explain it to a 5 year old, I'm going to say I do mapping or do Google Maps with some kind of fancy data behind.

Andrea Ciaffi

That's great. One of the things I really like to do is to ask people that have a career like you, what's the best piece of advice you ever received in your career because you have worked a very big firm. You've worked at Arup as a data leader.

Andrea Ciaffi

You work at Oregon and and As for many people like you have transition to running your own business in that and other management. But what's the best piece of advice you have received in your career?

Josh Symonds

I'll probably received a couple pieces of advice. I think the the first one was from my boss so whilst I was at 15 years and and Andy lost my boss there. One time I remember we were, you know, setting up for a project, and he's like, his comment was just do Josh things.

Josh Symonds

And that was kind of for me, was like, OK, just be yourself and everything will fall into place, which is, which was quite a nice panel, quite a nice one. On the Arab team, I suppose again was Neil Mumford. And you know what?

Josh Symonds

I've probably put too much pressure on myself from a work aspect. He had a comment about, you know, perhaps a big old ship. And if you fall off, you know, it's gonna take a long time for that to come round to, to pick you up or they won't be picking you up.

Josh Symonds

And to focus on family, I think that was also a very kind of key thing. And actually one of the the key things that made me drive to, to start in my own organisation, to be able to focus on family more. And then I suppose finally.

Josh Symonds

My last one would be it's more of a somewhere I look up to or someone who's more of a mentor. Where's Daniel Meehan? He's a he's a north now. But, you know, starting my own business and kind of like trying to figure out what to do and how to kind of pursue things and stuff.

Josh Symonds

And and always do. Like, what would Daniel do? And because he's just like this, you know, relentless, kind of like always on the go, always wanting to meet people and and, you know, do things together with people. And I think that's quite a quite a nice.

Josh Symonds

Yeah, quite a nice three, I suppose. Pieces of advice or or, you know, mentoring that I look to.

Andrea Ciaffi

Now this is great and thanks for sharing this because I think it's very important. Like sometimes there are these gems or not to Nuggets of gold that people say and can sometimes shave your life and your career because now you start thinking in a slightly different way how that can be useful for everybody else as well.

Andrea Ciaffi

But thanks for sharing that. Before we start with the questions related to Dada, can you start, I I know that when I was thinking about who is the right person to talk about data, how the world, the people I know and then a few people in this in this industry.

Andrea Ciaffi

So Josh is probably my number one choice for this specific topic, but can you start telling us a little bit about your career journey and what led you to work in construction in the data space?

Josh Symonds

Yeah, so an interesting career. I've had a very varied career. I started off long, long time ago. I won't say when, but you could probably tell by the grey hairs. It was a while ago. Oh yeah, I started off at a survey firm, which was a survey firm for a few years.

Josh Symonds

When I first left school, did a little bit of extended council, and then shifted to Brisbane. So I was in in Cairns and shifted to Brisbane and started working in consultant engineering, where actually the first project in Brisbane where I met my wife.

Josh Symonds

And then kind of shifted from from smack, which is when I first came down into into Arab. And I think that's where things really kind of exploded for me from a learning and a career's perspective as that I was able to, you know, coming to to airport link.

Josh Symonds

Doing the CAD side, focusing on on the bus ways and things like that, but then I was able to expand out and explore things.

Josh Symonds

So like going back to my survey routes and looking at GIS and then the mapping aspect came into it and and understanding you know, consult engineering in the database and then that kind of made me move into that data leadership role you mentioned before as well.

Josh Symonds

So. Yeah, it's kind of it's. It's not one kind of linear archaeology graduate engineer, and I've always worked in civil engineering. It's kind of moved around a bit, which is, which is quite a nice. Yeah. Keeps it interesting, I suppose.

Andrea Ciaffi

Yeah, no, it's definitely interesting. And as I always say in this podcast, like, it's interesting to see how different people that participate to these digital transformation come from different angles to then find their own space to, to contribute because there is, there are so many skills that are needed for the digital transformation.

Andrea Ciaffi

It's not just one person or one skill set that can solve the problem for all of us.

Josh Symonds

Yes. Yeah. No, exactly.

Andrea Ciaffi

So changing gear and and talking about data finally. So we know that Rala Dada plays. A crucial role in the construction industry, and it's a critical asset. But why do you think that is becoming such a critical asset in the construction industry? What's its role?

Josh Symonds

I think it's really become and it's always been this, but I think it's becoming more and more prevalent is that it's it's the missing link. You know you've got design is is really ultra detailed and data-driven.

Josh Symonds

You know using using digital processes to to generate a model and to you know to design your say say a road along for example you know you've got all these different software platforms that come together and they all talk to each other and then it kind of disappears at the issue for construction stage it goes into construction and I think really kind of has been coming out.

Josh Symonds

You know there's little bits and pieces getting generated. Way through the rfis and non conformances and stuff, but not necessarily that kind of same level of detail documentation you've had during the design phase. And then you've got the asset management phase which is very much you know how you you maintain the asset for the rest of its life until it gets disposed of.

Josh Symonds

So there's two very, very big data-driven elements. But then there's always been that little bit of a gap in the middle. And I've, I've, you know, noticed definitely in the last probably 10 years there's been a change in industry and you know part of that's probably been.

Josh Symonds

Suppose a push from external forces from maybe from the asset management side or or probably a little bit of getting ridiculed by the design teams, that they're not necessarily generating that data, but.

Josh Symonds

It's it's really starting to change in in that space and I think that, you know, the documentation of going through because that is it is critical for the asset management side understanding why things have been designed a certain way is great, but then understanding why it's been built a certain way is also just as critical.

Josh Symonds

I think as how it's designed and very, very critical for the asset management phase. I think.

Andrea Ciaffi

Yeah, definitely. Then and definitely and we know that we also have a big a big important factor of the digital transformation is beam build information modelling and what do you think, how do you how do you think that the role of data has evolved for the beam workflows over the last, let's say 10 years or few years?

Josh Symonds

Yeah, yeah, I think there's probably there's definitely been a push back into the BIM space to get the data out of those models and have more detail in there. So it could be things like, you know, carbon like materiality and how you can kind of get that information out rather than just being well.

Josh Symonds

There's adoring this location. So it's kind of that layering up of of more information in the, in the models, in the BIM models, sorry, BIM models, the ATM machine thing there. But I think that. As part of that as well, like there's been the demand for having these really rich data rich bins coming out of the design and into construction.

Josh Symonds

But the generation of that data, you don't want to manually go and create those things. So I think that automation has really kind of helped push into this space as well.

Josh Symonds

So things like you know, having Revit families and stuff like that to automatically generate as much data as they can on the way through and it, you know, saves kind of data errors as well with people manually inputting things. But I think that's really.

Josh Symonds

It's really kind of enabled those models to come out in a lot more rich form, a lot more once a useful form, but a lot more usable data within those models coming out.

Andrea Ciaffi

Yeah, that's right. One of yeah, one of the difficult things to do when it's about data because we have so many different data sources, data types. It's related to data management and now for construction projects. So we need to collect the data we need to organise it.

Andrea Ciaffi

We need to maintain it in, in beam environment, but not only beam environments. And we are also still running a project. So what from your point of view, what are the key principles? Of effective data management for construction projects.

Josh Symonds

I think other than referring to like the usual kind of data standards like Dharma and stuff like that. I think that there would be a standard really like having a standard.

Josh Symonds

It doesn't necessarily have to be a good standard, it just has to be a standard way of structuring things and what that gives you is a basis that you can kind of manoeuvre from, you know, if you don't know where your data is or you don't know how it's structured, it's really hard to try and move it anywhere or do anything with it.

Josh Symonds

Whereas if you know you know, hey, this is what my data looks like, it might not be a good standard. But you know what? We know what it looks like and then we can we can update that we can manipulate and change it into something something else or.

Josh Symonds

You know, you might have a spreadsheet, for example, how do you take that spreadsheet and turn that into points on a map that then correlate to to where door frames are or windows are? You know, if you don't have it that that core data standardised, it makes it really, really hard.

Josh Symonds

So I think having standards and like metadata cracking of where that data's come from and how it's been generated and stuff like that is really really key to help with data management.

Andrea Ciaffi

Yeah, that's that's pretty nice to hear. And one of the participate to many conferences where people are talking about data and I found it was very interesting because then people start arguing about which standard we should use and then we they end up creating a new standard to rule them all.

Andrea Ciaffi

So we have another another additional standard. So we end up having so many standards at the end. But yeah, I agree. I think a standard will make a big difference. Yes.

Josh Symonds

Yeah. Again. Yeah. And that's exactly right. You know, it doesn't have to be the perfect standard and there is no standard for all them all, but if you've got somewhere that you know where you are, then you've you know you've got direction you can, you know where you are with that.

Andrea Ciaffi

Yeah, definitely. And like, you know, I know you've worked on large projects or all sorts of projects, small, medium, large. And one of the challenges with large projects is how do we handle large and complex data sets. So what do you think, how, how do you think we can manage the challenges of?

Andrea Ciaffi

Working with large and complex datasets.

Josh Symonds

I suppose one of the struggles I've really had with with working with large projects, especially linear projects where you've got something with a lot of, you know, link and in complexity is is the level of detail. You know there's there's things like, you know, can have log 500 and model for example, which is great.

Josh Symonds

However, if you have a log 500, you might, you might only need 20 vertices to document a bolt or or make a a pole look nice visually and also accurately. But then sometimes you know. If you move data around in different different software packages, there used to be an old old issue with with.

Josh Symonds

When you'd draught a 2D layout of an intersection in CAD, and then you pull it into 12 D and it would do this weird thing where they make an extra vertices like every 10 mills, which doesn't seem like much. But then when you're talking like A10 kilometre project, that's a lot of extra vertices and that's just on one line string.

Josh Symonds

So it makes things really heavy and really tricky to work with. So I think that one of the things I would say I've definitely, definitely had a lot of pain in dealing with is having over verticied geometry. That is not so to say it.

Josh Symonds

If you have geometry that's got too many vertices, it's impossible to work with. You can't manipulate it. You can't work with it in an easy way, whereas if you kind of have the the amount that you need to do what you need to do, I think is quite a key thing to do there.

Josh Symonds

And if there is extra ones trying to thin them down for the next person as well. I think that's definitely one of the the biggest struggles that I've had is is the geometric side. The other side with working large complex data sets is.

Josh Symonds

They're not all compatible. You might have data that's coming from from 12 B and you're trying to combine it with Nic file and you're trying to combine that with, you know, two flow data that's coming up, getting those into a standard format is is not impossible.

Josh Symonds

GIS has really been useful for that because everything generally has a special context, and you can have your own schemas behind that. But I think that that's the the interplay or the interoperability between different software packages probably is the second biggest pain point I've had with with large and complex projects.

Andrea Ciaffi

Yeah, definitely, because this projects these tools have been developed by different vendors, different companies with different goals in mind. And of course now they have created these ecosystem of products that tend like struggle when they need to talk to each other. But what are some tools or platform that you think are most effective for streamlining this data management?

Josh Symonds

Not, not not. Dive into like my favourite piece of software, but it is is I use fmea lot which is feature manipulation engine. It's actually a GIS piece of software but it can read and write so like 400 or 500 different formats and it's incredibly useful for that interop piece.

Josh Symonds

And I I say I'll use it probably. Well definitely use it everyday if not multiple times a day just to pull data from 1 space, reformat it, restructure it or whatever that might be. It's incredibly useful for streamlining data.

Josh Symonds

You can have things like automations run, so you could have all these different common data environments and they have that all synchronised into one database and have that run on a schedule basis or based off of, you know, when things get versioned differently.

Josh Symonds

And because they can handle some of these geometry types, it can do that vertically, thinning that I was mentioning before. It can validate against a standard. You know, your data that's coming in, making sure your data types are correct and stuff like that.

Josh Symonds

So it's it's probably my go to for it. I know there's other products out there as well, like as you're data factories and stuff like that. But I think definitely, definitely FMA is probably the most flexible that I've I've used.

Andrea Ciaffi

Yeah. Thanks for showing that. And so now we have collected the data. We've tried to manage it. We found platform to share it and now it's it's it's time to turn this road out into actionable insight. So how do we report what we have and how do we make decision based on data.

Andrea Ciaffi

So data-driven decision for example. But from your point of view, how can project teams leverage the data reporting, data reporting to improvement decision making?

Josh Symonds

I think the the biggest thing and I kind of touched on this briefly before, is is pulling all the different disparate data sets into one location or at least having them cross compatible the amount of times that there's been a design review and someone's going, oh, I've output this into this format.

Josh Symonds

Can you load that up into into your software package and and review it or do again class cheques and stuff like that or getting stuff into IC to do a class check on it? Those kind of things I think are really, really useful to have.

Josh Symonds

So getting everything into a consistent. At least they didn't do a consistent area that people can actually do some reporting on.

Josh Symonds

I suppose combining different pieces of of software as well, so like using power BI or or something like dashboards and stuff like that to pull all the data together, but then also kind of do the dash section of it and kind of go OK well, where's this or how's the interplay between the constraints that we've got on site to what we've designed stuff like that is really,

Josh Symonds

really useful. But it's also quite tricky as well to try and get those all to play nicely, as we're saying before, about ladder products. And now there's there's an ecosystem over here, there's an ecosystem over there and they don't really want to talk together because they don't want you to to, you know, swap platforms.

Josh Symonds

Easily. That's why it is. But yeah, I think definitely getting everything into the one spaces is is a good way to to start reporting things back out. The other one good thing I've actually come across as well is if you run a workshop with with the design team to get, you know, kind of coming back to the design team.

Josh Symonds

But yeah, OK, cool. Well, here's here's all the data that we've got. If you wanted to go and look at you, do your design cheques or if you wanted to go and do understand where all the RFIS are, all the all the NC Rs are for our project.

Josh Symonds

And if that relates specifically to a design package, it is actually asking those questions if you, you know, to do your job day-to-day, how do you go about it? How? What kind of things have you always wanted to know but never actually been able to, to understand or whatever it might be?

Josh Symonds

That's always been a very powerful question to kind of sit everyone at a moment, go through.

Andrea Ciaffi

No, that's great. And I know that you work on many projects. Hopefully I'm not asking to reveal any confidential information, but can you share an example where data? The reporting an actual direct impact on project project success.

Josh Symonds

I did, actually. I unfortunately didn't get off the ground, which it really bums me out. It's one of those projects that you just like. It would have been an amazing thing to work on, but was was Auckland Light Rail? They're working on there was, you know, I'd done some earlier phases in different cranches of work around, like the utilities and stuff like that in 2017.

Josh Symonds

And it kind of carried through to more recent time where I actually worked on the master data management plan and we'll kind of, you know, cobbling things together and trying to get, you know, as much information as we could about all those data models.

Josh Symonds

And someone from the carbon team reached out. And she's like, oh, you're getting all this, materiality all this really rich information from from the BIM, can you can you give me some data back out of that around the volumes and what material that's made out of and stuff like that?

Josh Symonds

And really had this, it actually was quite a quite a cool experience because it's like it's not only just like how we want a carbon report, what our design is doing and how that's going to carry through construction and and etcetera. But it was also.

Josh Symonds

Like this really kind of cool design where you kind of go. OK, cool. We've got we've got. Design model.

Josh Symonds

This is how much the carbon is for option A, but if we go to option B, the carbon is going to be an extra 200 million tonnes or you know 200 million tonnes, 200 tonnes because we've, you know, made the the the tunnel long or we've shifted the pool or whatever that might be.

Josh Symonds

And that was kind of a direct correlation to you know we've moved this change, this here's the impacts of that. And I think that really had a big for me it was it was nice to see.

Josh Symonds

Something that seems so simple for me in the back end, I'm just grabbing data from here and I'm putting it into a database to actually see the impacts that they would have on the design team. So that I can understand why you know why this doing this is having such a big impact on, you know, carbon of, of all things.

Josh Symonds

So yeah.

Andrea Ciaffi

Yeah. Yeah. Thanks for showing that. One of the key concerns that we have when we talk about data is of course, information security. We see that all the time. Big companies get hacked by Haggar all around the world, and that is stolen. It's a disaster when that happens, there is reputational damage, there is data loss.

Andrea Ciaffi

It's bad for the users that now they're data being compromised. So we really want to protect sensitive data when it comes down to construction data and we want. So we want to ensure cybersecurity. But from your point of view, what are the? Biggest information security risks for the construction industry.

Josh Symonds

It's an. It's an interesting one and it really depends on on the scale of the project. I mean, you know, resi projects like building a house and stuff like that, probably not as big of an issue as it is for like massive linear projects.

Josh Symonds

I know there is the Sochi act, so the security wasn't security critical information. I think it is or something like it's all critical infrastructure. Sorry, I'll probably butcher that. I'm going to get pins for that. But but you know, if you're doing Resia project, it's less of an issue than it is on say, if you're building something for, you know, copper STR.

Josh Symonds

Massive transmission, electrical grid, waterframe planes, those kind of things, I would say. My rule of thumb would be do the zero trust approach which is very much a cybersecurity's, you know, rule of thumb I suppose you just do that, you know, zero trust at the first get go and then add access as people need it.

Josh Symonds

Do you actually need that to do your do your job? Yes. Then we can give you access to that I think that's probably the biggest risk. Is almost. We go the other way. Sometimes we're starting a project. We go, oh, we'll give everyone access to all this data.

Josh Symonds

Don't actually need it all for their day-to-day job. I think going through the zero risk angle is probably the biggest thing, and there's a lot of. There's a lot of accidental things as well.

Josh Symonds

There's probably another big security risk where, you know I need to get this across to, you know, Joe Bloggs, he needs something urgently to do something out on site and maybe instead of going through DOC control, you'll just attach it on an e-mail and send it through.

Josh Symonds

I've accidentally misspelt his e-mail address and now it's gone to someone completely different. Just little mistakes like that. The the bigger risks, I would say, you know, aside from the the very much normal cyber security of you know making sure you're clicking on the right links and stuff like that.

Andrea Ciaffi

It's often the final user that is the weak link when it comes down to the security. Yeah, I remember speaking with you, you mentioned ask you this question that this is a full profession within itself.

Andrea Ciaffi

So just maybe do do you want to explain like that there is a role for people manage the data, but there's also maybe another role for people to make sure that that is secure. Can you elaborate a little bit on that?

Josh Symonds

Yeah. Yeah. So it's it's very much a profession in itself. Like, you know, cybersecurity is its own. It's an industry on its own, really. I think from a, from a people protecting their own, their own personal data, not to mention project data, is again that kind of like, does someone actually need this, how if they got access to this information, how could people use this maliciously?

Josh Symonds

Sometimes it's not an easy way to kind of go about things. As in, you know the thought process of like oh, it's only just information about, you know, a power pole or it's only information about where a substation is. That you know on itself might be benign information, but when you combine it with other things could actually be used maliciously.

Josh Symonds

So I think that that's something to to always have at the back of our minds. And actually it's like we always think about, you know, project data is in the physical form like the digital form, which should say in the physical form, the digital form.

Josh Symonds

But then there is also things to be mindful of as like conversations that you have. You know in public or you might be, you know, walking. You might be catching the bus or talking to phone to someone on the bus.

Josh Symonds

You might have a competent conversation without realising it and giving away key information, again without realising it, but it can be related to a construction project as well, so it's always good to be to be mindful of your environment.

Andrea Ciaffi

Yeah, definitely. And are there a specific standard or certifications that teams should follow to ensure data protection?

Josh Symonds

There's a really, really good section in the Dharma or the data Management Association. We've got a book of knowledge that they've released many, many years ago, and they're like it's it's Chapter 7. See the six or seven? Just did it around data security and it covers things like, you know, data protection, confidentiality and normalisation.

Josh Symonds

And it touches a little bit into things like, you know, role based security and stuff like that. I think that's something very, very useful. If you're working on critical infrastructure, you know the Sochi Act is is very key for that. And on a personal level, you know you can brush up on the same the GDPR.

Josh Symonds

So the the Data Privacy Act as well, just to kind of understand what you know, what you should and shouldn't be doing. I suppose it's probably the biggest part of that and there's quite a lot of that is also in like micro credentials for universities, for example, they've run a lot of those little, you know, here's a six week course on data security.

Josh Symonds

Here's a six week course on you know. Had to do make and bin model. He is you know, so there's a few things there we could, yeah definitely look into if anyone's interested.

Andrea Ciaffi

Very important aspect is the legal aspect and the regulatory considerations around data. Now we're not here to provide legal advice to anyone, so speak with a lawyer and also you're not a lawyer, but you'll be in this space long enough that you probably have many conversation around the legal aspect of of data.

Andrea Ciaffi

So what are the legal consideration that considerations that teams should keep in mind when they're managing project data?

Josh Symonds

I think probably ownership and this is probably not necessarily during a project. It's more the kind of post project piece of work. I remember working with via at our many, many years ago that you know he's often doing his own thing as well.

Josh Symonds

But we've worked in and around like who owns data and how we, how can we keep using that data going forward. A lot of organisations are generating a lot of data from these these projects you know you you get gigs or terabytes worth of data.

Josh Symonds

Related to a project and it's good to kind of use that, whether it's for training AI models or just keeping it for future record. You know I need AI need a detail that I used on this project and you go and grab it out of that out of the coffers for that.

Josh Symonds

But a lot of a lot of organisations don't actually realise that they don't technically own that data.

Josh Symonds

The the client owns that data and I think that's that's probably something to really consider is, you know, maybe change your contract a little bit so that you can actually retain that data and use it going forward for your own purposes rather than having to destroy it as as per the contract.

Josh Symonds

But I think that that's that's also something that's a that's a big risk for a lot of organisations, is not necessarily realising that you can't just keep reusing that same data again and again. You don't actually own it but say. Another one would be, I suppose it's huge risk, but where your data lives as well so you know at the moment everyone's shifting to the cloud.

Josh Symonds

You know that's probably more of a COVID push, but everyone's putting stuff in the cloud is that that's great. Where in the cloud does that live? Is it in Australia? Can I specify where the cloud that lives? Then it lives in Australia. You know, there's the many years ago we were working around, you know, doing some stuff in in the US and and.

Josh Symonds

You know, having some service based in the US, but having can Canadian projects hosted on it? But the Homeland Act, Homeland Security Act in the US means that they can effectively skim through your data to make sure that you're not, you know, capturing anything malicious or anything like that, or any any threats to national security.

Josh Symonds

But that's obviously in breach of other data requirements or other data laws that we are, you know, bound to as part of a project. So we end up building another service stack in Canada and then having one of the America's just to kind of split the two.

Josh Symonds

So I think that that's that's also a very key thing to to be mindful of is where is your data residing, you know and is that a a driver for the project, you know defence projects can't have anything on overseas, so that's yeah, read the read the fine print and understand where your data lives.

Josh Symonds

It's it's a very good thing.

Andrea Ciaffi

That residency is a big is a big topic and if you know this sometimes when you start using a a tool you find that the free version of that tool doesn't let you choose where the data is stored, so your Basel giving your data.

Andrea Ciaffi

To to the company that is that owns the product you are using and then only then you realise that your transition in the so sometime maybe you are working on a project you say OK, it's free, I'll use this product, but always think about where does the data leave and.

Andrea Ciaffi

Is this aligned with my company? Needs very good point.

Josh Symonds

Yeah. And I suppose that comes back into that ownership piece too. Like, you know, sometimes if you're signing up to use a freeware or a shareware or something like that, not necessarily you own the data. So you might put data into that system, you know, chat, GP TS are a perfect one.

Josh Symonds

You put data into that ecosystem. It's then owned. They use that as part of their training data set. So they own that data going forward. So it's always something to be mindful of and check the fine print, all those things.

Andrea Ciaffi

Yeah. Yeah, another interesting evolution I've seen recently is that it's around the father. You pick one residence or one country for your dollar centre to be hosted in and then legislation can change later. So a country can always change the legislation. So it's not this is not a decision that you make today and it's true forever.

Andrea Ciaffi

It's something you always need to keep in mind. And depending on how.

Josh Symonds

Yes.

Andrea Ciaffi

The politics evolved in that specific country. Things can change.

Josh Symonds

Yeah, correct. And it and it has that has impact on the project as well. So you know, so if you're running a project in Australia and you have your, your data set was based in Europe, there's latency issues there of of you know pulling and pushing data is going to take a lot longer.

Josh Symonds

So there's a lot of considerations especially in that residency space.

Andrea Ciaffi

The performance aspect as well, definitely we will talk about this. The other residency forever and I know the Pro and 2nd related to where the data actually leads. Yeah, definitely. But how can from your point of view teams ensure that accountability and ownership of project data?

Andrea Ciaffi

What can they do?

Josh Symonds

I think one of the things that's I've seen in the past that's really worked well is is having data custodians.

Josh Symonds

So you have someone who is in charge of that data and it could be a team lead, you know, it could be the flooding team lead or it could be the electrical team lead that are the data custodian, although sometimes it doesn't need necessarily be that team lead, but you have someone who has ownership over the data, has that kind of I suppose they they.

Josh Symonds

They you know, parent from a better term, they become the the parent of that data and then make sure it's accurate. It's it's you know always as complete as possible and you know you're having that as a really, really powerful role because that means that you've got not only.

Josh Symonds

You're someone who kind of takes care of it, and someone who's actually passionate about it and and, you know, a little bit defensive or would make sure that it's getting used correctly. But that means that they're working with others as well in that similar role and kind of going, oh, hey, what data have you got?

Josh Symonds

I've got this data that's really OK. Let's share. And they might also share knowledge as well. So around, you know, different approaches to to, you know, processing for example, I think that that, that custodian. Kind of environment is really, really really useful.

Josh Symonds

Maybe mapping out that kind of ties back into the security aspect as having having a custodian and then understanding who's the users of all your data as well, because then that'll help you set your permissions and understanding how they're using that, and then you might go oh actually we need to capture extra data so that, you know, Joe from accounts can pull this in or or, you know someone from from their,

Josh Symonds

you know, from the carbon team. So they can have stuff or might add an extra column in and that really will help grow out the datasets. Quite rapidly.

Andrea Ciaffi

For sure. And we know that data management, it's not an easy task. There are so many aspects as we're we're talking about this a second ago, the residency, the the top of the data you you capture like the fact that the legislation can change.

Andrea Ciaffi

So there is lots going on. But I want to focus on the challenges related to data management and so we can help people avoid some most of the common mistakes. But from your point of view, what are the most common mistakes that you've seen?

Andrea Ciaffi

Produce team makes with data management.

Josh Symonds

I think there's probably there's a couple. There's a lot, really.

Andrea Ciaffi

There's a lot.

Josh Symonds

I can talk for days just on that. Really. No, I think I think one of the things that gets done a lot of is collecting absolutely everything, like just this stockpile stuff, really. There's obviously storage implications of that. And you know, it's it's not necessarily a good way of running things.

Josh Symonds

You know, why are you collecting, you know, drainage information or electrical information in a civil model when there's a separate model specifically for that one, you know? There needs to be a kind of a calculated approach to that, and that comes back to the custodian piece of like, you know, oh, you're owning this data.

Josh Symonds

OK, well, I don't need to have that in my models, et cetera, and which is really helpful. I think the other really, really big thing. And I've had a horrific experience with it in the past is is data lineage. So understanding where the data's come from.

Josh Symonds

So having that that metadata, you know the data about data. So how is it created? What software did they use, what accuracy is the data they put into it and. Yeah. What? What date was it created? Really. Really simple things like that, you know, mapping projections all all the above that is, is something that's not always collected.

Josh Symonds

And I worked on a project where we'd missed collecting or or documenting our our data lineage for the whole length of the project and then got to design delivery and then, OK, cool. Here's the data. Ah. But there's no like, no data lineage or no metadata attached to all this.

Josh Symonds

This data we've got. And we had unfortunately get someone to sit and manually input all of that for a couple of months. You know, retrospectively, because without having that information about where the data was created or how it was created, there's no trust you can you can't use that data and go, Oh yeah.

Josh Symonds

Now I've analysed and this is, you know, all the properties that were you're impacting or or analysed this and this is all of the you know I know there might be a chilly units got an issue and here's all the chillers that are associated that that are going to have a similar issue.

Josh Symonds

You don't know because it's that that data's got no trust associated to it you don't whether someone's accidentally copied and pasted it or. Or how it's been generated. So I think the the metadata and data lineage is a very key thing that. It's missed and it's it's it's actually quite a critical piece.

Andrea Ciaffi

Yeah, one of the key challenges that in general we have in the construction industry, this is not just related to the digital aspects, but in general is the speed and the fact that there is deadlines we need to meet specific deadlines, it's non negotiable.

Andrea Ciaffi

Other industry are a lot more flexible, but construction in this is not very flexible when it comes down to deadlines. So at the same time, we want to make sure that the data we collect is accurate. Everything has been done properly. So from your point of view, how can we balance this need for accuracy?

Josh Symonds

Incredibly sorry.

Andrea Ciaffi

With the pressure of a fast-paced project timeline.

Josh Symonds

Yeah, it's, it's, it's a. It's a genuine struggle. It's it's almost that kind of slow down to speed up if that if that makes sense. So you know as we talked about before like that geometric accuracy as well. You know if you're not pulling as much geometry through as many vertices through that can also help with getting things running a bit quicker or getting things out quicker.

Josh Symonds

I think the the in you know, speaking back to that red families aspect as well, like having things set up to be auto generated as much as possible. Umm, that's really really key.

Josh Symonds

Having things that you know you can just drop in and it's like, OK, cool, all the materiality, all of the, you know, sub elements of that are all generated automatically will really, really help with with getting things, you know, getting things generated quickly.

Josh Symonds

I think the the probably one of the big things is is tooling. So building a tool as as daft as it sounds sometimes going, I've created a tool that does this validation on this very bespoke and specific thing. Having that validation and having that error trapping has massive ramifications, especially when you're trying to turn things around quickly that you can just go OK, cool.

Josh Symonds

I'll throw it into the script and then I'll know what's wrong with it or what. What I need to change. I think that's really really. A really key one.

Josh Symonds

But yeah, I think there's a there's definitely A level of sometimes we need to slow down to speed up as yeah, as backwards as that sounds slowing down and making sure that we're following all our processes and procedures as much as possible.

Andrea Ciaffi

Actually, for sure and one of the key aspects when it's about data management is people and the fact that sometimes people don't have the right skills just because they've never been exposed to that specific challenge before, that working on a project. So from your point of view, what are the best ways to train teams to adopt A better, better data management practises?

Josh Symonds

One of the things that I've implemented in the past is like like I'm trying to think of the right temper. It's almost like rigid flexibility, if that makes sense, where like a naming standard. For example, having the ability that it's not all like pre mapped out and you've got to kind of go scrolling through 50 or 60 different things.

Josh Symonds

But having the ability to go, OK, brand naming standard when I'm start with naming standard, but it's like you know the date. Underscore we're only looking for 3A characters, and then three numerical characters, etcetera. And just like giving that flexibility in in naming things.

Josh Symonds

Like having kind of diving into that a little bit, but like having that ability to give a little bit of flexibility for the teams to kind of make their own decisions, because I have found and engineers in the worst for it, just putting that out there.

Josh Symonds

Now I'm going to get rid but is if you put something in that's too rigid and it doesn't do what they need to do, they will find a way around it. And I think that's also speaks to to data management. I mean that speaks to pretty much any.

Josh Symonds

Any digital skills like if you put something rigid in place, people will just walk around it or figure a way around it. So I think given that that flexibility, but I kind of controlled flexibility is is really quite quite useful. But I think that's the train teams.

Josh Symonds

I actually think probably getting teams together is is really quite useful. So you know, having the designers involved with talking to the asset management team, who'd you know might be the end custodian or the end steward of that of that, you know asset that's getting built is really quite useful because.

Josh Symonds

From a data management perspective, you got the asset team going. Hey, you know what'd be really cool and it'd be awesome to understand if how the paint was applied when they built this. You know, there's a standard that they did it, but why did they do this one pole slightly different?

Josh Symonds

That kind of conversation of like how you know RF is or NC Rs or stuff like that are captured and carry through, you know, could bear a lot of fruit.

Josh Symonds

So I think that there's, yeah, I've kind of meandered a little bit there, but there's there's quite a few different ways that you could probably have that, that better data management practises on projects and you know, kind of start with pulling people together and allowing people to to do their job without too many.

Josh Symonds

Restrictions guard rails guard rails.

Andrea Ciaffi

We calculate. Yeah, we are. We are close to the end, but I'm very, very keen to speak with you about the future trends in data. Now we know it's very hard to predict the future, but you've been in this space long enough that you may have learnt from the past and maybe have some visions or ideas of where this industry is going.

Andrea Ciaffi

So what do you see from your point of view, data management evolving over the next, let's say 5 to 10 years. So how far you can go basically?

Josh Symonds

OK. Oh, good. And I don't, I mean, obviously AI and machine learning are kind of at the at the forefront of a lot of that. I think it's it's probably going to there'll be a, you know, coming back to that reporting aspect that you mentioned about before.

Josh Symonds

I think there'll be a big change in that where it'll be like put all your data into one spot and get some AI to run across it and go. And then then then pull your insights out via that. I think that AI will have a significant impact on things like validation and potentially even populating missing values out of data sets as well, so you know.

Josh Symonds

We've installed 500 different things and it's like, oh, you've missed the this input on on this one value. Now it can actually probably help correct some of that, so it could help with that data accuracy, I suppose.

Josh Symonds

I don't think it's going to do data lineage for us unfortunately, but but I think that there's there's definitely been a change and I've I've noticed it in the last kind of, you know, 5 to 10 years, it's like gone from being 2D maps to now.

Josh Symonds

It's like 3D and and how we can get data integrating with each other. The digital twin space I can see that's exploding as well, so. There's definitely that kind of bringing together of data, combining it up into something like a digital twin.

Josh Symonds

And you know the Holy Grail for me would be having a digital twin with AI across the top of it and going, hey, I've got an issue with a tap of this type for a building. And it goes, hey, here's the five floors that have got that same tap installed.

Josh Symonds

You might consider changing these or, hey, you know, there's three of those taps were installed by subby one, and then two was installed by Subby 2. The two that you or one of them that you've had failed is from installed by this subby.

Josh Symonds

So maybe you just only have to check that other tap. So. There's. Yeah, digital twigs and AI. Suppose this is probably the key things that I can see is coming out as the the big winners. But even the way data stored is is, you know, changing prolifically as well.

Josh Symonds

You know it's. It's gone from being, you know, hard discs, servers on something, you know, pressed in to now it's all in the cloud denounce Geo redundant. You know you can have data replicated everywhere. So, you know, data lakes like houses, you know, structured data, all that kind of stuff.

Josh Symonds

So I think there's. Yeah, there's a lot changing this space. It's probably just who can stand on top of the quicksand for the longest, for the nicest, wait a second.

Andrea Ciaffi

Yeah, yeah. One of the one of the aspects that I am very excited about from the data and the AI point of view is the fact that we have, we have built large data sets over time and we ended up creating so all of these data sets.

Andrea Ciaffi

And one of the challenges was around how do we actually access the data, how do we actually see what's going on and how do we. Understand what to do about after we've seen what's what's in there. And I feel that. We have built lots of applications and what AI can probably do.

Andrea Ciaffi

That is my opinion, but we'll see what happens. AI can probably enable the creation of a new a new layer of applications that will enable people to access the data in a different way in ways that probably we don't know yet. That's an interesting evolution.

Josh Symonds

Yeah, it's kind of scary one though as well. Like, I mean, if you if you, you know, reflect back to, you know, the early 2000s, it was like what's an iPad most now? It's like everyone's got an iPad and and you can load models up on that.

Josh Symonds

It's like, what's the next evolution of that? It's it's quite, you know, it's it's scary and exciting at the same time. Like it's it's scary because I'm like, I feel like such an old man sometimes looking at all this new tech. But then.

Andrea Ciaffi

Yeah. Yeah.

Josh Symonds

Other times it's so exciting because you're like, oh, imagine remember, we did this project, you know, five years ago, we did it this way. If we had this technology back then, oh, my God, it'd be so much different. Like, that's really. That's the kind of exciting part, but then it's also stressful because I feel very old.

Josh Symonds

But at the same time. But I think, yeah, I think there's it's, it's quite an exciting time for data. It's getting valued a lot more, which is probably the the big thing, you know, start of the new world, all that kind of stuff.

Josh Symonds

But I think that there's been that realisation that, you know, it is actually quite a valuable, a valuable piece. One thing I should actually add as well that I'm really quite excited about is with. New data technologies and new ways of processing and and extracting data out.

Josh Symonds

Is all those old projects that we ran, you know, years and years and years ago where the, the methods or the approaches weren't as data rich as what they are now. Can we go back in time and grab those back out and now pull them into you know, can we generate a digital twin off of the last 30 years of data?

Josh Symonds

You know that we've got or you know for like main roads for example, you know, you look at all of the all the road plans that they've had over the last, you know. 100 years or whatever. Can you pull those out and get an understanding of how you know how defects are becoming a thing because of how the payments been created or whatever that might be?

Josh Symonds

I think that's that's a pretty cool. That'd be really cool tech to do.

Andrea Ciaffi

Oh, that's great. And I, Josh, I I think I could speak about data with you for the rest of the day. It's exciting that. Thank you very much for making the time to speak to me today. We are at the end of the episode, so if people want to reach out to you, where can they find you?

Andrea Ciaffi

I know you're also running your own business if you want to tell us a little bit more about that.

Josh Symonds

Yes. I mean, yeah, as we as I kind of alluded to at the start is is, yeah, start my own little little business if you want to reach out.

Josh Symonds

I'm on LinkedIn, we have a website, but you know it's not Ultra Ultra detailed but and yeah, you know kind of I suppose started my little little business to to go around and keep working with a lot of people that I've I've worked with before.

Josh Symonds

And kind of do more of that exciting stuff that I used to really love doing when I was at, you know, Caraport or Oracle. So yeah, I'm keen to keen to work with anyone who's got some kind of interesting, interesting data issues.

Andrea Ciaffi

Thank you very much, Josh. Thank you.