This week’s show is all about how Data Trusts and Project Data Analytics can help us deliver construction projects more efficiently.

I sat down with Martin Paver, CEO of Projecting Success and the Founder of the Project Data Analytics Community in the UK. 

Martin has over 30 years of experience in project and programme management across a variety of industries and is a Fellow of the Association for Project Managers (APM). 

In this episode, we’ll be answering some big question like: 

  • How can you built a community of over 4000 members all driving towards positive industry change? 
  • How can we maximise the impact derived from industry hackathons? 
  • What is a data trust? And how will data trusts impact the future of our industry? 

Interview:

Links:

Intro music by Vidian

Full transcript

Note: most of this transcript has been generated using an artificial intelligence algorithm and so it’s accuracy cannot be 100% verified.

This week’s show comes from London, I sat down with Martin Paver, CEO of Project Success and the founder of the Project Data Analytics community in the U. K.

He’s got over 30 years experience in project & Programme management across a variety of industries is a Fellow of the Association for Project Managers.

In this episode will be answering big questions like: How can you build a community of over 4000 members all driving towards positive industry change, how can we maximised the impact derived from our industry hackathons like the ones Martin produces and what is a data trust and and how they impact the future of our industry?

As you hear in the episode, Martin is incredibly passionate about the topic and driving positive changes in the industry, and it’s clear that the momentum is continuing to build. So without further ado, let’s dive straight into the interview with Martin from projecting Success.

Welcome to the Future Distributed podcast. Joined here today with Martin Paver the CEO of projecting success and also the founder of the Project Data and Analytics Meetup Group.

So thank you very much for joining us Martin.

Thank you will for the opportunity. It’s great to talk to you.

We’ve known each other for a while. I think that is the first time we’ve actually met, and that’s super excited. Yeah, right looking forward to it.

So if you could get us going briefly describing how you go into planning in the world of construction.

Certainly it was about 30 years ago. I started off in project Management. I’ve worked on various different jobs. The biggest one I worked on was a nuclear construction project worth about a $1,000,000,000 team of 220 people.

[02:37]

So I’ve worked in construction, I’ve worked in Defence and the Oil and gas sector and things like IT as well. ? Things that ideas well, so I’ve worked on big portfolios as well. So quite a eclectic mix of projects really.

Yeah, so you worked in like you say it’s lots of different sectors. How do they differ in terms of innovation, and their thirst for change?

So I think it’s about the appetite for change, Really. So I find in the IT sector, for instance, they’re just really working it cutting edge because they’ve got to be, because it’s the sector and always pulling in the latest content. in construction, I think there’s pockets of it with drones, for instance, are looking at the adoption of some of those technologies. But in terms of the things we talked about, which is Project Data Analytics, it’s a lot slower to catch on. I think we start to change that. But it is hard work, and I think through the community we can start to change that and through things like the Hackathons as well. We start to get some hands on experience as well. It starts to change, so we’re getting there. But it’s no easy.

Yes, you’ve mentioned a lot of interesting stuff there for us to unpack. So you mentioned your business Projecting Success, do you want to just give an overview of what you’re trying to achieve?

Certainly, it started five years ago and the first 3 years of that was me have being a project manager doing day to day work, running big projects, running a PMO, bidding for contracts in Project Manager World. And I sold out of another consultancy that gave me a cash to do what it wants to do with my life, and what I want to do is to really push the boundaries on Data Analytics. So I started off looking into lessons learnt and I realised that we’re not learning from one project to the next. And it’s not human learning that’s going to solve that problem because we’ll really know what’s going to go wrong in a project. We’ve got a load of insights, but we’re really flat out, and we tend to make mistakes and those mistakes in retrospect are fairly foreseeable. And I think what we actually need is some data on it is going on where we can start to share the schedules of the past the compensation events that we’ve seen before the cost plans, the resourcing plans, design challenges, technical challenges, etcetera.

So it’s this interconnected network of data. And once we’ve got that interconnected network of data we can start to interrogate it and get these insights out, which means we’re truly learning from one project to the next.

It’s not a statement in a lessons learned register that never gets read again. I think what’s important about that as well Will is it’s not just about organisational learning, It’s about sector learning. So if you give a job for a hospital through to construction company number one, it gets delivered. You complete next project and it gets delivered by somebody else. All of that composite Learning from the first job is not necessarily transferred over to the 2nd one. There’s competitive pressures in there. There’s all sorts of pressures that what I think for me, it’s a case of working out which of those pressures are truly competitive and which of those pressures probably need to be relaxed a little bit. I would say this is all for the greater good of the industry. We can afford more hospitals, so we get more construction work out.

[05:40] Great, okay, so you mentioned a little bit about the hackathons that you run. Do you want to briefly describe what goes on at these hackathons? And what’s come out of them?

Yes, certainly is for me, what we need to do is to bring together a data scientist, data engineers, app developers, those sort of people with project delivery professionals. And that’s when the magic happens. If you fill the room with data scientists they’ll code away all day long, but not necessarily bring it back to a project context. If you fill the room with project managers, have loads of enthusiasm but it’s not deliverable because it’s not grounded in reality. We need to solve that problem, and we solve the problem by bringing them both together. So what we do is we run three events a year that Microsoft Reactor in London. We get about 100 and something people along for the weekend. It’s capped out about 120. What we do is we got some challenges, and those challenges are based upon industry data. So that being from Sir Robert McAlpine from the A14 team Cambridge, it’s a big road construction project and we’ve had data from from publicly available sources and from the IT sector as well.

What we then say about these challenges is let’s work up a minimum viable product which enables us to look at the art of the possible and see what we can solve. So is to train inspire people to show people This is not scary to show people what can be achieved in a couple of days. And it’s not even two days, because when you look at the mobilisation of the team, they all start the forming, storming Norman type we got presentations on the second day is probably only about 7-8 hours off contact effort, and I think inspiration for that, You know, if you look at the last hackathon and the winning team, Sir Robert McAlpine said to us, we’ve got loads of these site diaries got 20,000 photographs in them. We can’t index those photos. We can’t find things in those photos.

As a challenge, could you find reinforcing bar if we actually had two in those photographs, because at the moment we have to search through them all. So a team worked up a deep learning model, rotated the photographs by 90 degrees, tried again looked at some different training data, etcetera, and it had a 92% accuracy of finding reinforcement bar in those photographs, they won first prize. They all got some things like Amazon Echo’s each. And it was quite impressive feat, actually, and they did that in a really short period of time. And if it can do that with reinforcing bar, they can do it with almost anything. So we can then start to tag these photographs. We can find things. We can relate that back to the schedule, relate it back to compensation events etcetera.

And it’s that connected nature of the data set. It’s not the data by itself, it’s that connectivity in the dataset which starts to give you these insights.

[08:24] To play Devil’s Advocate for a little bit, even though I’m a big fan of Hackathons, done quite a few in my time. What I’ve found is that really good excitement, And then some people say it’s just an excuse to eat pizza and stay up late. And then everyone got time and it’s all about so is there anything you’re doing to kind of nurture these ideas from that initial spark to getting value?

There’s a few things, actually, So there’s one of the team which was Monte Carlo’s Flying Circus, the name of the team and they meet once every couple of weeks. So the team keeps their inspiration going, they keep crunching the data and they’re trying to get into advanced project Data Analytics as well.

And that team keeps going from hack to hack and they keep on iterating. It keep on getting better and better, so I think that’s great. We need a lot more of that because that brings the community on. It means it starts to live by itself. It’s not just us who’s always feeding it. That’s great. I think we need a lot more of that. It’s the second thing is we were just working open innovation proposal at the moment, which is going to go into the construction transformation bid. Basically, there were saying, If we can use the hack a thon as a primer to reaching to a load of this data that we’ve got across the construction sector, what insights can we get from that data, and can we put some innovation money against it, which then takes it from an M V P into production of solution and there’s a lot of start ups we hope which will be created as a consequence of that will really start to drive it forward.

So I think that’s pretty exciting. If we do win that, I think that will start to transform these Hackathons from being something you turn up and eat a bit of pizza to. It could really change somebody’s life.

[10:10] Yeah, definitely convinced me. Yeah, and I wish you all the best. The more of these ideas we can bring into fruition the better. Uh, okay, So maybe on a related note, the project Data Analytics, the meet ups going from strength to strength, this seems like with over 3000 members in London?

It’s pretty close to 4000 members now in total were just about to start in Leeds and up in Aberdeen as well on in Terms of Aberdeen were working with the oil and gas technology centre to Mobilise it there as well. So they’re going to be a monthly events and if you get those going, we’re going to give it a push to get to 10,000 members by the end of next year, which would be great. You know, that’s a real force for good. And it’s not just bringing people together come and listen to eat pizza. We want to bring people together to really start to change the face of project delivery.

Things have not really changed in my career alast 30 years. You’ve got a few more tools, different risk Register, a different schedule whatever, but it’s fundamentally the same, I think we need to flip it upside down. Now, if you look at the stats on project delivery and there was a talk, which was done a few months ago with Alex Budzier from Oxford Said Business School. Alex said, And he’s done a load of research on this, he’s got something like 11,000 projects in his data set. And he said the probability of a mega project being on time on budget and on benefits is 0.5%.

And for me, that is a shocking indictment of our profession.

If you look at HS2 more recently, it’s just gone over budget by 20 billion. It sounds like a small number when you type it into a computer with all the zeros on, the end starts to bring it home. So that’s a massive number, massive overrun again and with Crossrail and we could go on and on and on. We need to change that. This situation is for me, Need some transformation applied to it, and that’s what we need to be doing.

And I’d like to be helping to drive some of that thinking and to really transform the profession. So that’s what excites me about it.

[12:16] Yeah, I’ve seen the calibre of the speakers you have and they’re super interesting talks. So if you’re in London or Leeds or Aberdeen, then definitely get on the Meetup page and Sign up for the next one. Building on that, you seem to be successful in building this out. What would you say is the key really to building on that community, finding a place that people want to come and contribute and learn.

I think it’s the niche. You know, there’s lots off meetups associated with data. There’s lots of Meetups associated with artificial intelligence. The institutions do meetups on project management, for instance. It’s for me. It’s the next big thing in project management will be really transformational and it’s bringing together those two different sets of people who start to spark off each other, start to learn from each other, and for me, to speak to people in project management and they can see this thing coming. And they think this is really going to change my job. It’s something that you got to engage with. I think if you don’t engage with it, you’re probably gonna be automated.

There’s gonna be some of these jobs, I think some of the project management jobs with a lot of people focus will endure, the ones that more process focussed, I think will get automated up to varying extents. So I think it’s something that people need to engage with. I think that’s really exciting. It’s not scary. It’s real exciting is what we can do now is instead of doing the boring work of filling in the Risk Registers and doing the schedules and all that sort of stuff, we can now look into the data set or this massive data set associated with Data Trust, for instance, on Bring Some Insights Out, which we never had before. And that’s the transformational capability in it. Really.

It’s interesting you mention about the nearly 4000 members, do you have any data on what the makeup that group is? What kind of backgrounds people have?

It’s quite difficult because with Meetup you can sign up with anybody so I could sign up as Joe Bloggs aand I put in a random e-mail address and that’s all you get. So it’s a limitation of the Meetup portal itself in terms of Meetup stats, we do do a Menti-meter session and about 60% of people engage in that.

We tend to get between a 50/50 split between project managers and or project professionals through to data scientists and data professionals. And in some cases it’s more Project Focus that goes down to 30% data. If it’s more data focussed, that goes to 30% project professionals. But there’s a real community to it as well, and in terms of that community, there’s probably 50 people we regularly see, and the rest of them come and go depending upon locations and topics. And we got the stats on that as well. And number of people who signed up to multiple events so we can run all the analytics on that.

[15:01] I wish you all the best for it, I think it’s really good for the community. So thanks for setting it up and hopefully good things come out in the future. So really want to dive in there on something you mentioned called a Data Trust. So for anyone that’s listening that doesn’t know what this concept of Data Trust, can you briefly outline what that means and what impact it could have on on the AEC sector?

Certainly. So in 2017 I think it was government wrote its artificial intelligence strategy and of part of that, it says we need to get more into data trusts. And he didn’t really know what data trust was at the time. And it’s been working with the Open Data Institute, and its put some definitions around that. There’s a load of definitions now. It’s starting to fragment a bit from data transferred to Data Commons, etcetera. And it’s, um, pedantry without, you know, quite different terms is generally the same thing.

It’s about the openness and the amount of data that you federate versus centralised. So with the data trust, what we’re looking at doing, basically it is to say to people if you share your schedule data, cost data, project profit data, even; some really sensitive data. If we can start to pool that data, and it’s not sharing, sharing is the wrong word. It’s got to be pooling the data. What we can then do is to get some experts involved. We can get some startups involved and say, Can you work up an algorithm that starts to predict the probability on a job? We can look at various factors, and if we change those factors get, then improves profitability.

We can look at ways we can look at all these different things because we’ve got a big enough data set. So if we just shared that data, everybody’s going to redact certain things that get anonymous it and it’s going to be useless this data set. So that’s not gonna work. It’s got to be a securely protected. So what we do is we put this data into something called a graph database, which is where we can connect all this data together.

So it’s not just a big data set our schedule data, a big data set of cost data, a big dataset of outcome data. It’s all joined together, and once it’s all joined together from one project, to the next project to the next project. That’s when you can query through the data, you can start to get these insights out of the data. So I think that’s, um, benchmarking in there. There’s some predictive analytics in there that’s assistance to decision making. So what would say to people is if you’re a constructor and You’ve got data to share. If we put that into this data trust and we can hen pool it with lots of other people’s data and the trustee, so the person with the data, can then say with that data, I’ve got right of veto. Who sees my data? And in What form? So it might be the raw data, it might be GDPR compliant data, where we’ve written the names out, maybe psdueonimised data, which is where you take the core data out but it’s then referenced back to an encryption key, almost so you can then de encrypt the data later and totally anonymised data, which is where you can’t cross-associate it later.

So once we got all that data, we could do some really, really cool things with it.

[18:14] Okay so that sounds super interesting. What kind of uptake have seen from industry at the moment, it seems like something that people might shy away from the beginning. What’s the reaction been like?

There’s 2 or 3 camps, I think, actually, so there’s camps of Well, this is never gonna work. You know, it’s the biggest excuse is actually the data’s rubbish. So if the data’s rubbish, what’s the point of sharing it now… That’s the biggest reason to start to pool this data is that if project professionals are making decisions based upon rubbish data, then we should all be sacked surely, because what we’re producing is crap, so that needs to change. And the only way we can change it really is to demonstrate that utility gap between the data, which is available today and the data we need in the future to do some really clever things. So I think that’s the first stepping stone.

So it’s to make sure that we’ve got data that is usable. It’s got some utility with it, so that’s the biggest blocker I think. And those people who just don’t believe it, you know, they think, well, this has never going to happen, it’s too hard, and I think we’re starting to disprove that now. We’ve been doing some work with oil and gas sectors, are we Oil and Gas Authority, the regulator for the North Sea, and worked up the documentation sets associated the rules. We’ve got technical architecture, so it’s all doable protecting this data, at a data component level is possible, and we can give permission to different people etc. So it’s all possible. We just need to go and do it now. One of the thought leaders on this is Sir Robert McAlpine, it’s a chap in there called Gareth Parkes, another chap called Grant Findlay as well, and they’ve seen leading a lot of thinking on this across the industry, and they’re strong supporters of it. They’ve been coming to the HACKATHON, and they’ve been sharing data with the hackathon, openly sharing data, which years ago wouldn’t have been possible. So they demonstrated. You can do this and sky doesn’t fall in when you do it, you know there’s not a lot of consequences of doing it. It just moves the industry on, so they’re saying, let’s start to change the industry.

And their chief executive wrote out, so Paul Hamer wrote out about three weeks ago to 15 other chief executives across the industry and said, But would you like to join us for a meeting. I would like to explore industry engagement with this data trust, and it’s basically a question off: Would you like to get on the bus? If you want to get on the bus, you want to sat at the front, driving the bus and trying to steer where that it’s just going to go? Or do you want to sit in the back of the bus and just throw some data in every now and again, so that meeting’s going to start to probe that, And to explore that once we get this going, I think it’s going to get a head of steam. I want to get some clients on board as well. So once we get things like NHS, Department for Education on board as well, I think that’s when things start to pick up. So I see this is start to snowball, and just getting to the point where we got chief execs writing to chief execs is a massive achievement. It’s taken us a long time to get here, but it’s a few inspirational people who’s getting us there.

[21:00] Yeah, that sounds really interesting. Just thinking this through in my head, I guess if this is all successful, and lots of companies sign up, we could build a really accurate machine learning models, I’m just struggling to see what that looks like. Now every company has the same prediction capability. So how are they pricing their job?

I don’t think it is because if you got a cost plan, for instance, you might say that you can see my data to create some cost planning tools, but in terms of discounts I’ve got, I might just take that discount out right So I might not share things until I’ve seen the development of the APP. So it’s a case of this trust being trustworthy, and it’s gonna take away your commercial edge. You won’t deal with it. So it’s a balance between commercial edge off sharing your data. So that’s development of new ups and new capabilities that’s going to make you more profitable versus giving away all your trade secrets.

There’s a lot of people who say to me, well my data has got a load of intellectual property in it, and it doesn’t really. Tradecraft is very difficult to capture in data, you know. So your schedule, for instance, is that really going to tell you the way that you’re constructing walls. He’s just more about benchmarking, rather than technical know how?

[22:14] Okay, that was really exciting future, but we’re slowly running out of time. So I have to wrap up and I want to finish by talking through through four quick-fire questions, 30 seconds to a minute on each.

If you had to change one thing about the UK infrastructure engineering sector overnight, what would it be?

It’s this data thing, actually, it’s to recognise we’ve got a massive data ploom coming out the back of a project and some of that is asset data. We fixed that problem through BIM, or we start to fix it through BIM, and the rest of that exhaust ploom associated with project delivery data just dissipates. We’ve got to change that. Once we start to change that, I think there’s gonna be loads of things in that data which we’ve not anticipated yet. And I think it’s gonna change the way we deliver projects.

[23:00] What book can you recommend that you think everyone should be reading?

I think that’s quite tricky one. I’ve been reading loads of data books recently. So data strategy, A data transformation. There’s no books yet about the data vs Project Divide. Sometimes I do post a lot of blog’s online, but it’s just a bandwidth to do it really. But the technology fallacy. I thought was really good book as well and it’s not just about the Tech, it’s about the people. So I thought I was quite good. But yeah and it’s being recommended to me by somebody from the community as well. And he posted something saying, Yeah, that’s good.

[23:42] Okay, so things are going well for projecting success. So what are your goals for the next three years for company and also the Data Analytics community?

In terms of the company, people said to me, What’s my business plan? And it’s really difficult to work out a business plan that says, You know, in 18 months time I’m going to eight clients doing X y Zed because it’s a very emerging market. We’re creating a market at the moment, which is to say to people, Let’s start to leverage this exhaust plume of data and some people want to get on the bus, and some people don’t want to get on the bus. That takes a lot of time to explore that you’ve got to develop and change the thought processes. So for me, I think my biggest target is let’s mobilise this community that’s really start to get the data trust going out of that will spin a raft of opportunity for loads of people. I’m hoping that projecting success is right in the middle of that. Bringing thought leadership on from thought leadership comes opportunity.

Okay, finally. So, as you know, I’m about to step into the Nordics, Copenhagen and potentially further afield as well. And you must have a wide network of people all around the world, Who’s that One person I definitely go and meet?

The person who’s really inspired me and really pushed me, is just down the road from me – a chap called Dave Snowden. Yeah, and I brought him along because I have paid thousands of pounds to go and listen to him previously and his thinking has changed my thinking more than anybody else. And it’s the way we manage complexity. We tend to think of the construction programme is that you need to schedule everything out and really hold the plan. And things move, you know, environmental conditions that might be or a constructor going out of business or whatever… things move all the time. It’s the way you manage that complexity and he brings a new way of thinking into it. And it’s working out which parts of projects are fairly simple. Which one’s a sort of complex and evolving. And once we’ve got that, if that’s what starts to change our project delivery.

Yeah, and if you haven’t seen it and check out the video of Dave Snowden his talk at the Project Data Analytics meetup – that was particularly interesting. Where can people find you online if they want to follow what we’re doing?

So find me on my LinkedIn Page Martin Paver and find us on ProjectDataAnalytics.uk and ProjectingSuccess.co.uk and the Data Trust is at DataTrust.construction


Will Needham

Travelling the world talking to the people shaping our future.

0 Comments

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.