Episode 1: The Synergy Between Safety Technologies and Industrial Efficiency and Productivity Transcript
Maurizio Pilu (02:36.774)
Good day from London. My name is Maurizio Pilu and I’m the founder and managing director of SafetyTech Accelerator. I would like to welcome you to this new episode of our Insight Series. Right, so at SafetyTech Accelerator, we work with dozens of corporates, technology companies, and partners on very complex challenges in safety critical industries, including safety and risk. So.
Maurizio Pilu (03:03.61)
One of the hottest topics that we discuss with partners at webinars, events, but also people approach us and in day-to-day conversation with our partners is, how can we build better business cases for investment in safety tech? This is coming up all the time. So today we would like to unpack this issue a bit more and try to answer a few questions. What is the link between industrial safety tech and the bread and butter of business cases?
Maurizio Pilu (03:34.29)
efficiency and productivity.
Maurizio Pilu (03:38.118)
Are these links common? So are we finding these efficiencies in adoption of safety tech across the spectrum of applications? Or are they present in just few use cases? Or is it a derail in the detail? So could it be that it’s about how it is done? So maybe you apply safety tech in one way and no benefits. You apply it in another way, bang, and you see a lot of benefits. So.
Maurizio Pilu (04:06.014)
These are the questions we are trying to answer today. And we are trying to answer them with Shelley Copsey, founder and chief executive of FYLD Welcome, Shelley.
Shelley (04:16.719)
Hello, Maurizio Thank you for having me today.
Maurizio Pilu (04:19.614)
Pleasure, pleasure. Shelley, why don’t we start about you? So could you tell us a little bit about your background? But also how you came to found FYLD And I’m particularly interested as to whether it was a light bulb moment. One day you wake up and say, I’m going to do this. Or it was a particular insight you developed as you were doing your previous business. Or is it a bit more organic? So tell us about how you came to found this company.
Shelley (04:48.891)
Yeah.
Shelley (04:50.055)
Background really quickly, I started out life as a chartered accountant. So when we get to how do you make the numbers add up, you can only guess that I like to go into the detail of how they do add up. But I spent many years working with major industrial companies, frequently in the infrastructure and real estate sector and looking at financial performance, you know, often in a mergers and acquisitions style scenario where everybody’s trying to figure out how they get that little bit more to the bottom line, you know, to justify a very different purchase price.
Shelley (05:20.149)
I did that for many years and then I was actually just fortunate enough to be asked to take a commercial leadership role at the consultancy I was working in looking at how digital and data technologies would transform their world and you know began to follow a path that I found really interesting looking at kind of the intersection of artificial intelligence you know the built world whether it was real estate infrastructure and then to me the core of you know what we’re very good at FYLD is how do you begin to get the human to use these technologies to let them operate in a
Shelley (05:49.909)
really different way. That led me to the UK five, six years ago where I led a geospatial company. And then more recently about four years ago into the leadership of FYLD When you kind of ask the genesis of FYLD I can’t actually lay credit to being that person. And I think it was actually a quite young person that went out into the field one day. And we were working at that point in time with Ontario Teachers Pension Plan, one of the world’s biggest infrastructure investors,
Shelley (06:20.129)
which is Scotia Gas Networks, SGN, they serve about six and a half million customers over here in the gas sector, were very much doing a project looking at the safety, efficiency and quality of fieldwork operations. SGN wanted to look at how they could raise all three up together and I began to find this really interesting insight that in a data-led world, how could we shift away from a historic view of, oh we can invest in safety but that’s going to slow the job down
Shelley (06:50.089)
you know, oh, maybe safety will go up, maybe quality will, but the costs are gonna blow out. And this much more modern way of thinking of actually, we can raise all three together and really, really transform our business. And you know, there was quite a light bulb moment, but there was a much cleverer person than me I like to think about that said, paper’s not gonna fix any of these problems. You know, in 2020, about 2020, 2019, how can we use video? How can we use photos? How can we use WhatsApp style
Maurizio Pilu (07:08.79)
haha
Shelley (07:19.849)
a messaging to really transform industrial work sites and lift all of those three factors together and that’s really kind of the genesis of how FYLD began.
Maurizio Pilu (07:30.346)
Fascinating. I really like this story. So it was a mix between light bulb and insight, and also maybe determination to fix a problem, which is sometimes in Genesis or many great. Yeah, yeah, yeah. Absolutely, absolutely. So while we are at it, could you just summarize very quickly what FYLD does? I mean, give us maybe an example of how a FYLD worker would use FYLD on a typical day.
Shelley (07:38.875)
Yes, determination was absolutely key.
Shelley (07:54.747)
Yes, so FYLD absolutely focuses on the execution of field work in the infrastructure sector.
Shelley (08:00.091)
If you think you walk out the front of your house and maybe you smell a bit of gas leaking, maybe you’ve got a water pipe that’s got a burst, or maybe you’re working on a construction site. A field worker opens the FYLD app and from the moment they start their day to when the job shuts down, they collect information in FYLD. We analyse that information in real time. We help remote managers understand where they should focus their attention next to safely deliver the day’s operations. Very importantly, and regardless of whether it’s
Shelley (08:29.845)
safety, productivity, or quality, we help people stop sampling, we help them stop going site to site to site, and we kind of help them begin to say, where should I focus my attention next? You know, after three, four years of commercial operation, we’ve also been to a good half a million work sites. Every work site, we get, you know, 10, 15 pieces of really neat unstructured data. So we’ve got a stack of data to begin to build those predictions of how days will play out to fundamentally change.
Shelley (08:59.785)
our customers operational models.
Maurizio Pilu (09:03.958)
Thank you very much for that. Thank you. I think we are going to come back to this in the rest of the episode. So let’s dive into the questions for today. Right? So your website gave some statistics about some of the gains of using the platform and so on. Actually, you’re making them quite explicit with the percentages and so on. So my question will be…
Maurizio Pilu (09:33.854)
How have you managed, how has FYLD managed, to tell the story of safety, improved safety and so on? But also, how did you manage to tell the story and quantify the story about efficiency and productivity gains?
Shelley (09:53.079)
I think when you begin working with an application that’s aimed at safety, the most important thing we recognise from day one is a six-week proof of value will give you lead indicators of change, but you actually need to do a much longer proof of value early on in your customer or journey to prove out longitudinally what would happen. And this is something I think I take forward as a lesson as an entrepreneur. We were very blessed to work with SGM for 12 months
Shelley (10:23.493)
I say safety story, but actually like it’s a safety truth. We proved out with data over 12 months, the impact that we could have on each category of incident and injury. So that when we go to the market and we claim that we drive down incidents and injuries by 20%, it’s backable by data. And I think the most important thing that I learned and have learned on this journey is that safety piece, it can’t be a story, it can’t be marketing spin. These are the lives of humans as people leaving the house knowing they should come home safe.
Shelley (10:53.073)
you need to make sure that your product works. So I think that we were incredibly lucky to have that 12 month period where SGN and us really collaborated to understand at a deep level, which gave us the first step into market. I think though then, you know, equally for us is that we often know that safety buyers don’t have that bigger budget. So we’ve kind of learned that we can get people in the safety world that will understand fundamentally how FYLD is going to change.
Shelley (11:22.613)
behaviours on the work site, right? And they can get their heads around this very, very quickly. But we likewise, and I think this is again, very important for anybody commercializing into infrastructure sectors like we do, margins are thin. People cannot afford to invest in products that have a risk of cost blowout as a result of paying for the product or what the product recommends. And similarly, what we do very carefully with all of our customers is quantify what is that productivity gain
Shelley (11:52.753)
in market time to value six weeks and it’ll be 8 to 12 percent, we can actually give them the actionable plan that underpins it. What I do find though increasingly is that the maturity of the market, it goes up and up and up every year right, you see that the bar go higher. I think we’re finding more and more operations people now when we speak to them, they’re not interested in actually doing any form of project that could have safety risk. They actually want to see from the get-go how are you going to hold safety at the same or more importantly
Shelley (12:22.553)
how are you going to positively impact to it as we make shifts to drive our operational model and increase productivity and efficiency. That’s to me been actually a really pleasant part of the journey over the last four years seeing that maturity in the market that trade-offs shouldn’t be okay and that we can raise all together.
Maurizio Pilu (12:41.683)
That’s brilliant. Just to bring it to life a bit more, so can you give us an example of behavioral change that is directly linked to increase in efficiency? So how would you measure that? Maybe through an example, maybe some statistics that you could share with us.
Shelley (12:57.491)
Yeah. Let me give you the simplest one that we’ve managed to do, which is a risk assessment. Right.
Shelley (13:04.323)
We all know the best way to keep a worker safe at the start of the day is to get them to really, really in their heads, understand the risks that are in front of them and have them so far as is possible and put control measures against those risks. Right. And there’s so much science that sits behind where can you eliminate risk, et cetera, et cetera. In the infrastructure sector, you know, maybe prefab is becoming more common. A lot of the work we do, utilities, they’re digging up pipes, right. Robotics is nowhere near replacing the humans. So we can’t eliminate.
Shelley (13:34.197)
continue to drive behavioural change. You know, a simple example of where we get the two brought to life at FYLD, when someone arrives on site, we get them to take a video, right? We immediately get governance that every single person on every single job is doing the risk assessment where it counts, which is at the job site. The next thing that we get every single person to do is talk to us about the risk. Our AI does a wonderful job, but I will never get away from the power.
Shelley (14:01.591)
of the person saying, you know, that I’ve got a deep excavation and there’s like high voltage power, you know, the power of them speaking about it, vocalizing it, coming back to them as we heard this, are we correct? Can you confirm and have you controlled it? That behavioural shift is immense, but I think it’s enabled by real time data moving around and so you get really good governance over it. And positive coaching that remote managers can see this type of information and they can positively reinforce.
Shelley (14:31.465)
great risk assessment, really good to see this quality of work. But let’s go to the flip side onto productivity and what we see happen there if we get a good risk assessment done upfront. The first thing that we see is that operational managers are now getting a look into site as well in a way that they never can in a paper world, right? We can also use artificial intelligence that’s been run over this video to help alert those remote ops managers about jobs that look like abnormalities compared to a normal job. So they also know where to focus their attention.
Shelley (15:02.265)
When we begin to look at the outcomes of understanding the risks that are in front of me, I’m controlling them, I’m being made alert to things that I may not have seen, I’m being coached and someone external can get to site if they need to before something goes wrong. What we’ve seen this transfer into typical risk assessment, if it was taking 20-25 minutes on paper, we’ve got it down to three despite the 20% reduction in incidents and injuries. Absolute classic no trade-off
Shelley (15:31.385)
both of them, you know, two ships in the harbour rising together, but concurrently many of our customers are seeing uplifts in span of control, right? So because they’re no longer sampling sites, they’re not driving around needlessly, phone ringing all day, but they’re in control, they’re focusing where it counts, we see 25 to 40% uplifts in span of control because you no longer need so many remote managers attempting to cover your patch, you’ve got a focused number of managers where it counts.
Shelley (16:01.547)
that we see in the field.
Maurizio Pilu (16:05.206)
Thank you. Obviously, the decrease in time to do a risk assessment is such a clear-cut efficiency improvement. I just wondered, though, if there is a combination here. First of all, it doesn’t seem like there is one single thing that is driving that. It’s a combination of several things, almost like a cascade effect. But it is almost like a digitization dividend that you have by adopting your solution.
Shelley (16:12.892)
Yes.
Shelley (16:26.428)
Yes.
Maurizio Pilu (16:34.79)
Right? So we are moving from paper to digital. Then there is a way of capturing data and so forth, more visibility, more transparency. And then there is the specificity of the application. So it’s a combination of both. They kind of build on each other. And that’s an interesting pattern. Obviously, well, obviously, actually, I assume you’re gathering quite a lot of data about this improvement. So you’re building statistics on a client basis and probably across your client base.
Shelley (16:47.955)
Thanks.
Maurizio Pilu (17:03.946)
So you’re starting to have robust data, imagine about these improvements.
Shelley (17:04.144)
We are.
Shelley (17:06.939)
Yep, which you know lets us do a couple of things. It lets us begin, and this is where our research thrust is, predicting how a day will play out.
Shelley (17:15.151)
Right, so not only do I now in the moment understand the risk on a job site that somebody might be facing from a safety or operational perspective, but can we predict what will happen in three hours in five hours and stop it now? Right. And again, if I just look at this from a safety lens and, you know, we’ve spoken before that the OFGEM, the UK regulator for energy have funded us to about £1.6 million non dilutive because they see this as game changing.
Shelley (17:45.285)
site six hours in advance and let’s say now somebody intervenes and they stop that incident in its tracks right. Off-Gem funding safety plays is not their usual spot right like in the UK we’ve got the health and safety executive who would normally be the funder of that we’ve actually got OFGEM funding us because they know the implications to industry if we can get down the level of incidents right so you know as an example the utility sector in the UK
Shelley (18:15.105)
and injuries on sites right through to deaths. You know from our perspective if we can get that down by you know 20 30 percent there’s probably half a billion pounds in there of money that we can begin to go after and really shut down for industry. There’s the human and I don’t ever like putting human life into pounds I think is a fundamentally wrong concept but we do have to operate networks right. There’s the personal cost to people when things go wrong. There’s the
Shelley (18:45.025)
the cost of inspections and all of this is ultimately borne by the public in the price of their utility services. So you know it’s been really interesting to see OFGEM follow us and be interested in our perspective of how we can use all this data that we now have to you know predict how a day will unfold and enable us to help companies make sure more operations are smoothly completed without incidents.
Maurizio Pilu (19:09.27)
That’s amazing and congratulations on the funding for such an important project. I didn’t know there were so many accidents or incidents. I’m not sure which one. Maybe a bit of both. There’s quite many. It’s incredible. So I can see the hundreds of millions of pounds and dollars building up if this problem is addressed. It’s very interesting. So. And.
Shelley (19:31.291)
Yes.
Maurizio Pilu (19:36.562)
Let’s take a step back a little bit about what your product is doing. And I’m quite interested in trying to derive some general principles about how you manage to connect essentially a safety solution, but not just safety, as we said, it’s also digitalization with productivity. So how can we generalize your experience to other application use cases of SafetyTech?
Maurizio Pilu (20:03.894)
So what were the general principles? Is it about digitization, as we said before? Or is there something around improving the way an individual employee works? I mean, what are the general principles? If there are any, by the way, because sometimes they’re a bit difficult to find. The idea is that maybe to share some learnings to other entrepreneurs and also safety managers, operational managers that are trying to decide whether to invest in safety tech or not.
Shelley (20:30.715)
Well, I think that the biggest thing we learned in FYLD that we didn’t even maybe know early on was going to be so core to our success was how do you enable a customer, every utility, every construction company, they’re on a transformation journey, right? Everybody’s trying to figure out how to digitize so that they can lift all three of those things we’ve discussed. That’s kind of just like table stakes. But what we’re seeing concurrently is globally, there are labor shortages in our core industries. And in lots of sectors, there are skilled labor shortages.
Shelley (21:00.669)
We’re seeing inflation bite into everybody’s margins, right? It’s a world where most people are trying to do more with less. What do I see as kind of the generalisations that sit here? You need to attract better labour and you need to keep it and you need to make that labour more effective quicker. Everybody we talk to is trying to do that. And, you know, I think I’ve seen that throughout my career as I’ve worked from industry to industry to industry that’s digitised. The first thing that I think everybody can think about is AI can scare people and AI can make
Shelley (21:30.549)
people think that they’re going to lose jobs.
Shelley (21:33.107)
Right, lots of things we work in, that’s a long way off before that’s really happening at any level of pace. How do you use AI to coach and to enable people to be better and to be a driver of organisational culture change? It doesn’t need to be the core thrust of your product, but what we always think about is, could we give a manager a nudge that they should point out that someone’s done a really good job today and congratulate them? Why do I think about this? I think it’s near impossible to quantify the productivity uplift or the safety uplift
Maurizio Pilu (21:33.607)
It does.
Shelley (22:03.181)
doing that but I’m a firm believer in both upward spirals and downward spirals and if you can use you know AI as a coach in these types of dangerous environments to make people feel good right I think you begin to get an upward spiral of I’m going to work I feel valued I feel appreciated I feel like my work is seen I’m going to do a great job today conversely there’s a lot of good research that shows if I go to work and I’m disengaged and don’t feel happy I’m more likely to you know take some
Shelley (22:33.661)
And that’s not the human’s fault, that’s just human behaviour. So that’s my first one, like how can you use your technology to coach and guide and make people feel great and build upward spirals? You know, the second thing that I really think about a lot, and this has been, I guess, embedded in me since earlier in my career, when you’re working in consulting, you always have, you know, the four I’s it’s called, you never send anything out the door with just two I’s. Again, how can we use technology? It’s got to be positive. It can’t be someone sitting on your shoulder, you know, ticking, crossing the door, you
Shelley (23:03.041)
what you’re doing but as an enabler again how can we use technology to get more and more eyes on what is going on so we contribute positively to more work being done. I think that these are actually quite big generalisations that can be made when you think about technology to keep people safe but I just also think fundamentally there’s no other part of my career I’ve seen where people have been so open to making trade-offs. I think there’s a piece for safety managers and ops
Shelley (23:33.361)
where they’re beginning to have the education to understand that the technology can really enable them to achieve multiple goals together and that they shift away from a trade-off mentality. That’s been something that’s very nice and again I think there’s a piece for entrepreneurs to you know we largely adopt a challenger style sales at FYLD right. If you see people that still haven’t actually moved along that education curve what’s your willingness
Shelley (24:02.781)
just expand the minds of people to what is the art of the possible in 2024. When we do that we tend to find people massively receptive because perhaps they’re just not sitting with the same exposure to technology that we are. But these are always really bright people that we’re interacting with so it’s just having the courage to challenge some of the status quos and to show people what the difference could be if they thought a little differently about some of their business challenges.
Maurizio Pilu (24:30.922)
Fascinating. You have, I think, focused on the individuals, really. Individual as an agent of change, improving the way they feel about their work, augmenting how they work, but also the cultural aspect of digitization and innovation. I mean, from my experience, challenge number one of digitization is culture, the culture of an organization. And the funny thing is that there is no way back.
Shelley (24:54.908)
Yes.
Maurizio Pilu (24:59.318)
once you start, it just snowballs, right? And it’s one of those things that people never say, I need less data, I don’t need digital tools. It’s unstoppable once you get there. But people need to see the benefit, and you’re absolutely right, they shouldn’t feel threatened. They should see that as an opportunity to work better, maybe even have more time for more creative type of work. That’s very interesting, very interesting.
Shelley (25:21.171)
Wait.
Shelley (25:26.179)
It is. I think the other thing that we often think about is, I’m completely with you Mauricio, once you begin the transformation journey and you get CIOs, COOs hooked on the drug of data, they aren’t going back. But likewise in the infrastructure sector, what causes the failure of projects like a deployment of a new FYLD workforce management system? It is the workers, right? Is the workers rejecting the tool? Is the workers putting bad data into the tool? All of those types of things.
Shelley (25:56.153)
is such a focus for us at FYLD. People say you’re an AI company, those algorithms must cost a lot of money. I think that a modern company also says design, user engagement, reinforcing cultural values is equally important as that AI. We very much think about how do we get that FYLD worker engaged and feeling good through using a product. And I think that’s something else that we’re beginning to see the C-suite at our customers think about and be far more aware of like,
Shelley (26:26.113)
feeling good for the workers are they going to pull it and use it or is this going to be another product that has you know mass organ rejection we do with that by focusing very heavily on like a CSAT style score with the user let give them something that they love and that they pull to use in the FYLD.
Maurizio Pilu (26:44.374)
Really, really interesting. I mean, we also, when we talk about these matters with the dozens of entrepreneurs we work with, save some certain cases where really humans are not too involved. But yeah, human factors are so important and sometimes not addressed first. Sometimes it’s an afterthought. Yeah, let’s develop the tech. Let’s throw it over the fence and see what happens. But actually, designing with a user.
Shelley (27:04.143)
Yes.
Maurizio Pilu (27:10.794)
An improvement in the user’s life is probably one of the biggest, maybe, take away from this conversation. Thank you. Now, maybe we can zoom out a little bit further, and I would like to ask your views as an AI entrepreneur in general, as an AI entrepreneur. So obviously, it’s almost impossible to not see the kind of buzz that is going on around AI these days. I mean, in the past…
Shelley (27:10.909)
Yep.
Maurizio Pilu (27:39.754)
probably five years, it’s just exploded. And I just wanted to ask you, bearing in mind we are working in safety critical sectors, with sometimes fairly conservative sectors, I wanted to ask you what you feel about, how you feel about these bugs, right? You say, oh, you say, yeah, this is my kind of stuff, or is it, well, let’s be careful here. And so how do you feel about the bugs? But also I wanted to ask you,
Maurizio Pilu (28:08.358)
In your view, what’s the role of regulators? There is a lot of discussions these days about AI safety and safe AI and so forth in America, the UK, Europe. So as an entrepreneur in this FYLD, what are your thoughts about what’s going on out there?
Shelley (28:14.023)
Thanks.
Shelley (28:24.727)
I think two things, data security before we even get to ethical AI has got to be way up on your list and you’ve got to be investing in it because ultimately you are collecting a lot of personal data even if it’s some anonymized data, right? So that’s to me just table stakes for anybody who plays in these types of industries. Secondly, I think ethical AI is such an important concept in product design and how you think about building a company.
Shelley (28:54.421)
that we could probably build that are highly personalised. I very much am in favour of the regulations that we’re seeing come out of the EU, etc. That’s a bit like, what are the boundaries? I think that if you don’t have a regulator beginning to impose boundaries, the potential monetisation opportunities that come about as you build more and more datasets are huge. And I think ultimately what do we see in the world time and time again is that when those
Shelley (29:24.421)
go down a slippery slope right and at the end you realise how far you’ve come but on the way it’s little step after little step and you can lose awareness. So I think regulation is highly important although obviously very difficult as this FYLD of work emerges. But again like I think you can get clever in also how you think about some of this stuff in tech. I think maybe a little bit differently from productivity to safety. Some things that we think about from a safety perspective
Shelley (29:54.321)
in the moment, so say like an ephemeral style of tracking if it didn’t count. So if I just give you a little example and this isn’t something we’re doing but you know something that we sometimes talk about.
Shelley (30:04.559)
You get a lot of voice data in what you do at a company like FYLD. Could you actually understand if somebody sounds stressed today? You don’t want to do anything with that type of data, except for maybe alert a safety manager if there was a spike so that they could make an intervention. Right. I can see the trade-offs here from a safety perspective. If I could see in the moment that someone had a really stressed voice, there’s potentially a really nice intervention that stops somebody getting seriously heard on site.
Shelley (30:34.533)
you need people to opt into that style because it’s a personal choice to my mind. Do I want to be offered the chance to be kept safer and am I willing to you know forego some level of privacy in the meantime? And I think that these are difficult concepts for companies to grapple with but I think these are the most important things for us to grapple with as we begin to grow data companies that are related to safety, productivity and where there is so much human data involved.
Maurizio Pilu (31:03.83)
Thank you for this, the latest example you have given about stress measurements. We have done projects with clients that wanted to, with full consent of the staff, experiment with technologies. And for me, that’s an interesting one because regulators could give the guardrail about how to do it. But then there is the people, do you want to do it? Do I give consent for this? So
Maurizio Pilu (31:28.198)
Even regulators, they might give you maybe a guardrail, but it doesn’t mean it’s going to happen or people really will accept it. So the human factors aspect
Maurizio Pilu (31:37.507)
comes into play again, right? Always, always, it’s all about people, isn’t it? Fantastic, fantastic. Yeah, always. Amazing insight, Shelley, thank you very much. We could speak for hours about this. I think we are both passionate about AI and these topics, but that’s all we have time.
Shelley (31:37.507)
in a thousand days. Yeah, always.
Maurizio Pilu (31:55.766)
for today. So I would like to thank you very, very much for the insights you shared with us today. Thank you.
Maurizio Pilu (32:04.81)
Thank you.
Maurizio Pilu (32:07.467)
I would like to thank you for listening to this episode of Safety Tech Accelerators Insight. Please feel free to contact us if you would like to discuss today’s topic, or if you would like to discuss collaborations with us, or if you have any suggestions about topics we could cover in a future episode. From London, this is Maurizio Piru. Goodbye.