Seb (00:02) 

Hello and welcome to this episode of Safetytech Accelerator Insight Series where I’m going to be talking to Helen Balmforth from the Health and Safety Executive about their approach to sandboxing. My name is Seb Corby and I’m a Principal Consultant with Safetytech Accelerator. Helen, welcome. Good to have you here. Could I start by asking you to introduce yourself? 

 

and your team and what they do within the HSE. 

 

Helen Balmforth (00:34) 

Yeah, so hi Seb. Hello. So I’m the head of data analytics at the Health and Safety Executive and I work out of our Science and Research Centre which is at the laboratories in Derbyshire. So our work is all about really trying to understand how we can use data and new analytical techniques to improve health and safety performance. So we’re looking at trying to extract new insights and sort of do things differently. So I work across HSE 

 

but also out across into other government departments and also out into the private sector working with industry and tech companies. 

 

Seb (01:11) 

Brilliant. And could you talk about the objectives of your team specifically and what you’re trying to achieve? 

 

Helen Balmforth (01:18) 

Yeah, that’s fine. So we’re part of the science side of the organisation. So we’re very much coming at how we can use technologies and new techniques with data to really understand how we can push forward health and safety risk management, understanding the workplace of the future, that type of work. So we do work with our regulatory colleagues, but we’re coming from the innovation of the science side of things. 

 

So I work with a group of data scientists. We also work out with other partners. We’ve worked a lot with the Safetytech Accelerator, as you know. We’ve worked with other government departments. We’ve worked with academics. And we undertake projects that try and sort of tackle these sort of maybe bigger issues and try to bring that learning back into HSE so that as a regulator, we can learn and we can understand how we can improve as a regulator as well. 

 

Seb (02:07) 

When we talk about sandboxes, it’s quite a broad term. 

 

It involves research and testing of technology. And when did you decide that innovation sandboxes were a good tool for you to use in your team? 

 

Helen Balmforth (02:21) 

Yeah, so I think there is something around the term sandboxes. So we’d heard people talking about sandboxes, but for us, we wanted to see what we needed in HSE to be able to start to address some of the challenges that we will have going into the future and the workplace of the future. So for us, it wasn’t about suspending regulation or anything like that. Our sandboxes are about creating that collaborative environment to work together with all the different stakeholders that we need to. 

 

So it’s about bringing in industry, it’s about talking to tech companies, it’s working with like the private sector like yourself to really tackle issues together. So we very much want to take this approach because HSE, we need to maintain Great Britain’s great record as one of the safest places to work. That’s one of our missions. And to do that, we need to understand new risks that are coming into the workplace. We need to understand how new technologies can help manage 

 

current and new risks as well. And so we wanted to try and approach things a little bit differently and try and see if we could use a sandbox to do this. And we very much want to be an enabling regulator. So we want to help people innovate safely and doing this together collaboratively, much more upfront before risks or technologies are employed in the workplace and risks have manifested themselves. If we can work together. 

 

upfront we can sort of have a better understanding. So that’s why we thought about using a sandbox but it isn’t business as usual for HSE. We’re still trialling it through the innovation work that we’re doing in the Science and Research Centre. 

 

Seb (03:55) 

So the first sandbox that we worked on together was looking at barriers to entry within heavy industries like construction for several key technologies from machine vision to robotics and other forms of AI. Could you tell us a little bit about that sandbox in particular and possibly what are the learnings that the regulator took from that sandbox? 

 

Helen Balmforth (04:20) 

Yeah, no problem. So the first one was the Industrial Safety Tech Regulatory Sandbox. It was funded by the Regulator Pioneer Fund. So, and they sit in DCIT. And it was actually quite a short project. It was eight months. So it was, you know, all hands on deck. We were, you know, working at a pace, but we were trying to understand how new technologies might help us manage some of the health and safety risks that we see in the construction sector. Construction sector is still one of the most dangerous places to work. 

 

So we wanted to sort of try and do things differently. So we identified a set of risks that are real priorities for HSE. We also talked to industry stakeholders and we wanted to understand which risks were their greatest worry. And we came together and we actually, you know, very much have the same risks that we’re all interested in. But then we had to have a technology company that could come in and help us address those risks with a new solution or a bit of tech that they’ve got. 

 

And so we had to that tripartite approach to tackling this. So once we’d identified those challenge areas, the risks, the risks that we were interested in, we then went through the process, working with yourselves to narrow down the tech companies that could address those. And we ended up with five really, really nice use cases that were addressing different risks with different bits of technology. we had things like computer vision, 

 

and sonal working, looking at plant -person interface, which is still a big issue. We had technology looking at how we could integrate live data into risk assessments and working processes actually in the construction workplace. So it was really trying to push the barriers on some of the tech and how it was used. We didn’t necessarily think there would be trials. 

 

in that first sandbox because it was a really short period of time. The sandbox itself was only about 10 weeks. But actually, in the end, some of the companies did actually do some trials, which was a nice additional benefit. But what we were trying to do is more feasibility studies. So we published these and each of the feasibility studies has got a really nice in -depth review of how the technology was, you know, trialed the sorts of issues that came up, how we address some of the barriers to adoption. 

 

So we’ve got really a lot of detail in those reports. But just trying to summarise, we did review across those five projects that we undertook and we drew out themes. So we drew out themes around barriers to adoption of safety tech and then we were trying to address those. And the themes were largely around culture, technology and then financial sort of incentives to be able to adopt industrial safety tech. 

 

It was interesting that actually regulation didn’t come up as too much of a barrier. It did feature, but it wasn’t the main barrier to adoption. So we’ve got some really nice insights there and we’ve been able to take that learning and also how we ran that sandbox. It was really interesting to get the learning from how we interacted and how we all worked together. We’ve been able to take that into the second sandbox, which we’re running now. 

 

Seb (07:32) 

Yeah, it’s, it’s probably a bit of a fallacy that a regulatory sandbox is only about, like you say, suspending regulation or informing regulatory development. I think one of the things we found was that actually having the regulator in the room and being involved in the trials and the development of the technology was beneficial not only from, I guess, informing yourselves about possible changes, which 

 

like you’re saying, actually often regulation wasn’t the limiter in and of itself, but actually engaging with the regulator, the experts that are within your team also really benefited the technology companies. And I will say, I know one of the trials actually showed that the issue wasn’t the regulation, I guess, wasn’t up to date or was behind, but it was actually people’s ability to 

 

check their compliance within the development of robotic design. So one of the trials and feasibilities was around how do flexible robotic systems ensure that they remain compliant with safety regulation. And actually, the regulation was OK, but people’s ability to efficiently and effectively monitor their design against the regulation was the challenge. And I think certainly one of our 

 

Helen Balmforth (08:53) 

Yeah. 

 

It was interesting that one, wasn’t it? Yeah, because it showed, think it’s sort of like the system, it was a system wide issue really, wasn’t it? So the regulations were perfectly okay, but it was just the process of having to do all those compliance checks. And it was a real barrier to people, you know, being able to sort of think that they could actually adopt this in the workplace. So by doing that trial and sort of unpicking some of that, it was really good learning for us and industry. And as you know, we’re sort of still 

 

Seb (09:00) 

Exactly. 

 

Helen Balmforth (09:27) 

taking that forward with thinking about different systems that could allow easier checking of compliance at all those different stages when you’ve got something like, you know, robots or whatever working in the workplace, adaptable robotic systems. Yeah, no, it’s interesting. 

 

Seb (09:45) 

One of the other benefits that we saw from that sandbox was actually the closer working relationship between industry, like construction companies and the technology companies themselves. And I know we saw some technology companies actually kind of radically altering their propositions based on the feedback and insight they got from the sandbox, which was really positive. We’ve now taken that model of industry regulator. 

 

and technology engagement and we’re doing another sandbox between ourselves. Would you be able to introduce from the regulators perspective kind of what your objective is with this new sandbox? 

 

Helen Balmforth (10:29) 

Yeah, so this new sandbox is the Smarter Regulation sandbox. So this time we’re funded by the Government Office for Technology Transfer, who sit in DCIT as well. So again, this isn’t business as usual for HSE, we’re still doing it through our innovation and our science work. But this sandbox is running for 14 months. We’re looking at how we can make our regulations smarter and machine readable, and actually sort of opening up our data and our regulations. 

 

to be able to help innovators innovate and provide solutions to help improve compliance and health and safety performance. And if we can do it for health and safety regs, the model could be adapted for other regulatory areas as well. So it’s really nice. We’re working with the smarter regulatory directorate who sit in DBT, Department for Business and Trade. We’re working with other regulators as well who’ve got a, like a watching brief wanting to see how we as the HSE can sort of… 

 

make some progress with this. So it’s got quite wide stakeholder group and lots of buy -in. And what we’re doing is we’re trying to understand how we might need to provision our regs and our data in formats that can assist with this machine readability and being able to then allow innovators to take it and innovate and provide the solutions. 

 

So the role of the Smart Regulatory Directorate in this piece of work is that they’ve created something called the Open Regulatory Platform. And basically it’s a platform that could be used to provide sort of a one -stop shop for businesses to understand what regulations they need to comply with. But at the minute, it hasn’t really got a lot of content on it. At the moment, a business wanting to start up would have to go to all the different regulatory areas. It would have to know what it needs to know as well. So there’s not one place that they can go to. 

 

So we’re trying to test whether we can provision health and safety regulations through the open regulatory platform marked up in a way that allows them to be accessed by algorithms that can make it machine readable and innovators can take that and then go on and create solutions. And again, we’ve got five projects that we’ve got on the go, all looking at different sorts of ways of doing this, all looking at different sort of tech. 

 

actually accessing different bits of regs and different bits of data. But again, we’re hoping that we can get that nice view of what we need to do as a regulator to provision our data in a better way, how we can then pass that learning on to other regulators so they can do the same. Then also understanding the benefits that can be realised from access to smarter regulations, which is where, you know, we might be stimulating the economy, we might be making it easier for new businesses to start up. 

 

which is a game where a smarter regulatory directorate come in. So yes, it’s a nice project. Just taking it further, the learnings from the last one further into this one as well. 

 

Seb (13:23) 

Yeah. So I mean, I guess to put it into context, one of the projects that we’re looking at is how a large language model that currently ingests regulations from scraping PDF documents online, how the regulator could provision data directly into a language model that would possibly improve its performance or enable it to get access to other data. It sounds like one of the 

 

core threads running through the sandboxes is that it’s not just about you yourselves learning how to possibly adapt your regulations or become smarter, but there’s a big emphasis on, I guess, technology acceleration and growth and collaboration within the sector. Could you kind of talk a little bit about that side of it from your perspective, what you’re looking to achieve in the kind of the wider 

 

Helen Balmforth (14:12) 

Yeah. 

 

Seb (14:21) 

from a wider lens, I guess. 

 

Helen Balmforth (14:23) 

Yeah, so we very much know that new technology is out there, it’s been used in the workplace, we know people are scraping content from our website, we know their chatbots, but at the moment as a regulator, we want to be able to sort of understand and be forward thinking and proactive in how we can understand and manage and help others understand and manage the risks with respect to health and safety as this new technology comes into the workplace. 

 

So it’s very much about time to do this collaboratively. We can’t do it on our own. We need to work with industry. We need to work with the technology providers that are creating these tools. And if we do it upstream before they’re deployed and before they’re creating them, before they’re put in the workplace, we’ve got a much better environment to understand together and to learn and to actually create solutions that are really going to have an impact in the workplace. That’s what we’re trying to achieve. We’re trying to achieve. 

 

the management of health and safety risks, improving health and safety performance and making sure that more people don’t get injured or made ill by going to work. That’s ultimately what we’re trying to do. But we know that new technology can offer such potential to help improve that. So we’re really trying to look at the opportunities, trying to understand the barriers, trying to unpick those and help facilitate how things might be taken forward by industry. 

 

But we have to do that upstream. have to work with tech. We have to work with industry to do that. So I suppose that’s the bigger picture for us. 

 

Seb (15:59) 

From our perspective as a technology accelerator, we work with lot of industrial companies and we obviously work with a lot of technology companies. And there’s industries where technology has advanced and is coming to the mainstream. And there’s other industries where possibly things are flagging. And as we’ve said, it’s not just kind of regulations that are really preventing. 

 

technology uptake. And one of the interesting things about, I guess, your sandboxes is that you’re looking at it from a wider lens than just yourselves and tech or yourselves and industry. You’ve run these sandboxes from the perspective of all stakeholders. And to me, it seems to be quite unique, certainly from the kind of regulatory sandbox perspective, but actually, 

 

from the perspective of really kind of government innovators and industry innovating together. Is this something that other parts of government are doing? Have you seen other governments doing this or how unique do you think this approach is? 

 

Helen Balmforth (17:01) 

Yeah. 

 

Yeah, so we know that there are lots of sandboxes being run across government, but they’re all a little bit different. think what we’re trying to do, and it may be because health and safety regulations cut across so many different sectors, it’s so broad, but we’re trying to sort of understand, I suppose it’s the ecosystem. We do not just sit in isolation as a regulator. We sit in the UK PLC system, and it’s a really important point to sort of know that. 

 

Our regulations are goal -based, so we’re not prescriptive. Some other regulators have different approaches, but we’re not a prescriptive regulator. So we allow innovation, but we need people to understand and manage their risks as low as reasonably practical. That’s how the health and safety system works at the moment. So what we’re trying to do is we’re trying to understand as a regulator how technologies are sort of growing and the opportunities for the workplace. And then… 

 

how we as a regulator, what we need as a regulator to understand. There are still sort of bits around standards and sort of benchmarking the performance of some of these new technologies, but that’s getting into the regulatory space. And we’re not at that point at the moment. At the moment, we’re looking at it from an innovation point of view. So we’re looking at exploring that side of it with technology and industry. So I would say we’re fairly unique across government within the UK because 

 

We do have this large science division within HSE and we do have a remit to sort of undertake this type of sort of innovation and science activity. We’re also benefiting from the investment that governments made through the Regulator Pioneer Fund and through the Knowledge Assets Grant Fund, which is the CAGF system that got administered. So we’ve been beneficiaries of those funding schemes, which really have allowed us to accelerate what we’ve been able to do. 

 

And while we’re testing it on health and safety regs and we’re trialling it, you know, with our regulatory area and with HSE’s remit in mind, it could be picked up and sort of used. Some of the learnings that we’ve done could be found, could be used by others. So we have got interest from other regulators. We are talking to others. We know that people are sort of looking at what we’re doing and hoping, we’re hoping that we can sort of transfer some of that learning to others. So I think… 

 

I don’t know whether we’re completely unique, but we’re just a different way of doing this. And it’s really valuable for us to explore that in this different way as well. 

 

Seb (19:45) 

So quite often, one of the options that I guess the government department has is to go and do more of an academic study, maybe a kind of digital strategy, study on what should you be doing in terms of adapting your regulation? How should you be interacting with AI moving into the future? And you kind of get a really nice document at the end of it. 

 

And there’s possibly sometimes been a criticism, I’d say, of studies like that, because they say fantastic things, but then there’s no tangible outcome and step forward. So does this kind of represent an alternative mechanism for learning about things like AI and how you should be interacting with them? And is there something to be said for being extremely practical and industry driven? 

 

about these things. 

 

Helen Balmforth (20:47) 

Yeah, so I think that’s an interesting point because I do think we do a bit of both actually, Seb, to tell you the truth. But one thing that we’ve been able to do with these projects is that we’ve really been able to actually try implementation. We’ve actually done some doing rather than just the thinking and the producing the, you know, the documents and the approaches and the strategies. We have got those and we are developing those. But these projects have allowed us to push forward the other side of things as well. That isn’t to say that we’re just 

 

Cavalier, you know, racing ahead. We are doing this in a very staged way, in a framework. And you know, we’ve developed this framework, but that’s what’s really nice. That’s what’s the transferable bit. It’s almost like we’ve done the thinking, we’ve done the working out and the understanding. In many cases, we might have de -risked it for other regulators to adopt this approach. So we’ve got that framework, we’ve got criteria, we’ve got… 

 

sort of guardrails around what we do in these sandboxes, what we do in these studies. I think we’ve been able to sort of demonstrate quite nicely that we can still get really good results from doing this, perhaps more than you can from just producing a document. So actually engaging with industry, getting them excited about working with us, getting tech interested, allowing people to sort of come together and actually work collaboratively. 

 

I think has moved as a step much further forward than we would have done if we’d just been behind closed doors developing a strategy. We did hold an event at the end of the last sandbox, the Industrial Safety Tech sandbox. And that event was really well received and attended by industry. And that was really nice to hear those anecdotes from industry about how they’re encouraged to see the regulator sort of working in a little bit of a different way. 

 

and wanting to sort of do more of this in the future. So I think the approach that we’ve taken with this practical implementation of these sandboxes is definitely something to consider alongside the development of robust strategies and the rest of the research that we might undertake. 

 

Seb (22:55) 

So the final thing I want to talk about is, I guess, kind of what happens next. What are your plans for the future? And to lead on to that, one of the big things that we’ve both found is that there’s a really innate benefit from bringing technology companies, customers within the industry and the regulator together. And you kind of get all of these. 

 

benefits that possibly you didn’t foresee the adaptation of products and services in line with customer requirements, greater understanding of risks shared between the stakeholders. And I guess the question is, how do we make this more systematic? How do we get those benefits moving forward when we don’t have funding to run a sandbox, for example? 

 

Could you talk I guess from the regulator’s perspective on what you’d like to see happen moving forward? 

 

Helen Balmforth (24:00) 

Yeah, so we’ve got another six months or so in this sandbox that we’ve got at the moment and we’re looking forward to seeing what that actually delivers, which will be very, very interesting. But we don’t want it to end here. We’ve done two. We’ve got real momentum behind what we’re doing. We’ve got this different way of engaging with industry and really sort of having this collaborative environment. Not to say that we didn’t necessarily do that before, but we’ve got a really nice momentum behind this. So we’re really hoping that we can carry on doing some of this. So there’s probably two strands to this. There’s 

 

thinking about HSE, it would be nice if HSE could use sandboxing of some form as a form of intervention. So it’d be a really sort of upstream intervention way before you get down to an inspection as an intervention. So it’d be nice and we’re talking about including sandboxing in our intervention toolkit. Again, we’d need to bring it into HSE and think about how we would do this as business as usual, but we’re still going through that discussion. But then if you think about what we’ve done, 

 

in one regulatory space, there’s definitely scope to adapt this and move it into other regulatory spaces, especially with respect to safety regulation. know it’s been, you know, there’s a lot done in the financial sectors and they have maybe a different approach to sandboxing. But I think there’s learnings for the other safety regulators that, you know, we could help accelerate. So I think there’s something around a wider discussion within government about how we can keep the momentum behind this going. I think it’s possibly something that the new 

 

regulatory innovation office, for example, in DCIP might be interested in. Obviously, we don’t know the details of that yet. I do think there’s appetite within industry to keep this engagement going with us. And it may be something that we have conversations with industry groups. If they want to sort of keep this dialogue going and this way of working with us, there’s different options to do that. We are working on a network for innovation and regulatory safety tech, as you know, which is funded, a trial has been funded through Innovate, hasn’t it? 

 

We’ve got nice momentum behind that, which is just feeding into what we’re doing with the sandboxes. The sandboxes are just one sort of implementation of the types of things that might come out of that network. So if we can build that network further forward, we can have that open dialogue, we can start to address issues, then we might use a sandbox approach to address some of those issues, or we might use different approaches. But I do think there’s definitely momentum. We want to keep it going and we don’t want to just… 

 

stop when this particular project finishes. 

 

Seb (26:30) 

Could I guess push you one step further and some of the conversations we’ve been having is certainly the network for regulatory innovation and safety tech that we’re looking to kick off. But I guess there’s a bigger ambition within the UK to actually become a centre of excellence for health, safety and wellbeing innovation, certainly within heavy industries and possibly wider. I guess there’s not. 

 

a lot to talk about just now, but could you maybe touch on your ambition and where you possibly see that going? 

 

Helen Balmforth (27:06) 

Yeah. 

 

So we definitely have that ambition. feels like we’ve got something pretty unique here and we’ve got real momentum behind it. So we want to keep that going. But if you think about the whole OSH picture as a whole, rather than just particular sort of sectors or whatever, it’s absolutely vital that we understand how that contributes. So it feeds into the economy, it feeds into economic growth. So this is the innovation aspect of that. 

 

but it also feeds into that productivity discussion as well. Keeping people healthy and well in their workplace means they’re more productive. There’s a really important role for us to play in that. New technologies, new ways of doing things present new opportunities to facilitate and accelerate that productivity and that keeping people well and healthy and safe in their workplace. So I do think this isn’t going to go away. This is going to grow. We just need to find the right routes for us. 

 

But I think it’s really nice to think about the system as a whole. Often health and safety is just seen as a little bit like the poor orphan, you know, health and safety. And it doesn’t necessarily have budgets or it doesn’t have the high profile of some of the drivers that industry would have, but it is absolutely key to achieving some of those bigger ambitions. And we want to push that and prioritise that. And there’s conversations to be had across government, there’s conversations to be had across industry, but that’s our ambition to try and make sure that we can push that forward. 

 

Seb (28:35) 

Okay, Helen, thank you very much and no doubt see you soon. And to everyone listening, thank you for listening in. This has been Insight series by Safetytech Accelerator. 

 

Helen Balmforth (28:47) 

Thank you.