3089: Web Summit - Beam Introduces AI Subsea Robot for Offshore Wind Farm Inspections

[00:00:04] Web Summit 2024 is once again bringing together some of the brightest minds and industry leaders from around the globe.

[00:00:13] And today I'm thrilled to be speaking with Joe Tidball, VP and co-founder of a company called Beam.

[00:00:22] And they are a British deep tech company that is at the forefront of revolutionising offshore wind operations.

[00:00:29] And Beam is making waves in the renewable energy sector.

[00:00:34] And they're doing this by deploying AI driven robotic ships and autonomous underwater vehicles.

[00:00:41] Incredibly cool what they're doing here. Regular listeners will know one of the big motivations of this podcast

[00:00:46] is getting everybody to think differently about how every aspect of our life and world is impacted by technology.

[00:00:54] Even at the bottom of the ocean.

[00:00:56] So with their recent world first achievement of launching the first ever AI powered autonomous underwater vehicle for inspecting offshore structures.

[00:01:07] Beam is setting new standards for the industry.

[00:01:10] So I caught up with them at Web Summit this year in Lisbon to find out more.

[00:01:16] But enough from me. Let's get Joe onto the podcast now.

[00:01:21] So a massive warm welcome to the show.

[00:01:24] Can you tell everyone listening a little about who you are and what you do?

[00:01:28] Yeah. Hi Neil. Thanks for having me.

[00:01:30] I'm Joe Tidball, co-founder of Beam and executive VP of service innovation.

[00:01:34] It's my job to really embed technology into our operational side of the business.

[00:01:39] So taking everything that we're working on in autonomy, AI, computer vision and general improvements to our processes.

[00:01:46] I'm rolling that out into our operations.

[00:01:48] Fantastic. And before we learn more about Beam though, obviously we're here at Web Summit in Lisbon.

[00:01:55] So just to set the scene for our conversation today, can you offer a short intro to the offshore wind sector for a non-expert?

[00:02:02] Because I suspect we've got a lot of people listening, they're hearing about this stuff for the first time.

[00:02:05] Absolutely. Offshore wind is fascinating. I think it's one of the industries that's definitely growing massively, but it's often overlooked.

[00:02:14] Clearly it's offshore, so it's beyond your line of sight typically for if you're looking out to sea.

[00:02:19] So it's often a little bit forgotten, but there's a huge amount that's going on both in the UK, Europe and globally.

[00:02:26] It's a huge grain market and really pivotal in my opinion for moving us to a net zero future.

[00:02:34] Fantastic. And on this podcast, I always try and get people thinking a little bit differently about how technology impacts areas that maybe you don't associate with technology.

[00:02:43] So how has tech changed or is changing the offshore wind industry? What are you seeing here?

[00:02:49] Yeah, there's huge changes coming about in offshore wind through technology and we're leading the charge in a lot of those.

[00:02:55] But again, going back to the point of it's often out of sight and therefore out of mind, there's huge applications that aren't necessarily even being thought of.

[00:03:03] Sometimes I think in terms of solutions to problems you never knew existed, right?

[00:03:08] So there's a huge amount that we're working on to improve the rollout of offshore wind and its growth and its effectiveness.

[00:03:16] If we look at the full lifecycle of a wind farm, we think of it in four main phases.

[00:03:21] At the beginning, you have the site consenting where you're looking at actually deploying the wind farm.

[00:03:26] What does the seabed look like? What does it look like beneath the seabed? Are there any hazards?

[00:03:30] Are there any protected habitats that you might need to investigate?

[00:03:34] And we have a whole range of technologies that are helping to support that.

[00:03:38] So if you look at something like Explorer, our autonomous surface vessel, we have minimally crewed autonomous boats that are going out and collecting this data.

[00:03:46] We'd normally need 30 or 40 people on a large ship to go and get that data.

[00:03:50] We're doing that with two, which is obviously a massive improvement.

[00:03:54] Well, how do you do that?

[00:03:55] You know, how do you take 20, 28 people off a vessel and perform the same amount of work?

[00:04:00] But it really comes down to AI in the data processing pipelines.

[00:04:05] So taking data that would normally take nine weeks for a person to process, running AI algorithms on that and doing it in three hours instead.

[00:04:13] Clearly, that's making a massive difference on the data delivery turnaround.

[00:04:18] So if you look at technology in the offshore wind space, just told you that, you know, you've got a huge amount of data coming through that has AI potential and all sorts of other things going on too.

[00:04:27] Yeah. And I'm glad you mentioned AI, because obviously here at Web Summit and every tech conference is all anybody's talking about.

[00:04:33] And again, when we're talking about this sector, you don't automatically associate AI with that.

[00:04:38] So how widespread is AI, traditional and generative in this space that you're working in?

[00:04:44] Yeah. So I would say it's not particularly widespread at the moment at all.

[00:04:47] But as with anything, there's there's AI creeping into all facets of everything we do.

[00:04:51] But we really are embedding that into our day to day use cases.

[00:04:56] So if we look at traditional AI is really important to us, although it kind of traditional kind of sounds like old hat and, you know, maybe not as relevant.

[00:05:04] If you look at what we're trying to do in the inspection space of wind farms, we're determining whether a subsea structure is still able to support the weight that's on top of it.

[00:05:14] And the wind turbine is still able to to generate power safely and reliably.

[00:05:19] If you look at the traditional sense, what we want to do is determine that a defect is, in fact, a defect.

[00:05:24] How how critical is that defect? Are those deep defects happening on multiple assets across multiple fields?

[00:05:32] So traditional boring AI for us is really important because we want that guaranteed deliverable to the client.

[00:05:39] Generative AI, on the other hand, is potentially potentially problematic. Right.

[00:05:43] You're trying to generate new information from the data that you have already.

[00:05:47] And that could lead you down a path of, oh, well, you know, it looks like there could be a crack here, but let's move that out.

[00:05:53] And actually generative AI, maybe not quite as applicable for for what we're looking to do.

[00:05:58] So we want that really reliable, high throughput, very precise traditional type models.

[00:06:05] And it's so refreshing to hear you say that as well, because I think everywhere you go,

[00:06:08] AI is seen as that silver bullet to fix absolutely everything.

[00:06:12] Are there any other challenges that AI poses to you?

[00:06:15] And how are you bypassing them then?

[00:06:17] Because it's not as simple as just, hey, let's go, AI.

[00:06:20] There is a lot of problems to overcome.

[00:06:21] Yeah, exactly. So there's again, going back to what we're doing,

[00:06:24] we're running the full pipeline of projects for our customers.

[00:06:29] So we're deploying the vessels and the robots to go collect the data, processing it, reporting it,

[00:06:33] and then ultimately delivering that to our customers.

[00:06:35] So they need to be confident that what we're delivering to them is of the same or better quality than what would be done with people.

[00:06:42] Right. So we've got this 400 times quicker tool to be able to detect horizons in sub bottom profiler data,

[00:06:50] which is essentially the cake layers beneath the seabed.

[00:06:54] And it's all well and good doing that initial interpretation 400 times more quickly,

[00:06:58] but you still need a person to come in and actually verify that all of that is okay.

[00:07:02] Now we have tools as well that are speeding up that process.

[00:07:05] And in say, automated boulder picking,

[00:07:08] we're 57% quicker in our ability for a person to determine if a boulder is indeed a boulder and the size of that using AI.

[00:07:15] But the challenge is this one where you need to take your customers on the journey that okay,

[00:07:21] brilliant, you've got all these new WISI tools,

[00:07:23] but are they actually giving you the data that we need to the quality that we need it?

[00:07:28] And obviously we'll have people listening hearing about Beam for the very first time.

[00:07:33] We've set the scene about the industry now.

[00:07:35] Can you introduce everyone listening to Beam's technology and also what you're doing here

[00:07:39] and the kind of problems that you're fixing with your technology?

[00:07:42] Yeah, absolutely.

[00:07:43] What does Beam do?

[00:07:45] Well, I've already mentioned that we operate across the full lifecycle of a wind farm.

[00:07:49] And we split that into four main categories.

[00:07:51] So we have the marine site characterization phase,

[00:07:54] which is happening prior to the wind farm being built.

[00:07:56] Then you have construction where we support some of the smaller tasks in terms of site preparation.

[00:08:03] And if you imagine how expensive an offshore installation vessel is,

[00:08:07] you don't want your that big vessel doing some of the smaller tasks.

[00:08:10] So that's where we come in there.

[00:08:11] We then have the operations and maintenance phase of a wind farm.

[00:08:14] So what is the integrity of the assets?

[00:08:19] Is anything changing over time?

[00:08:21] But essentially just making sure that the wind farm is in a good operable state.

[00:08:24] And then we're a long way off it in offshore wind, but decommissioning will come eventually.

[00:08:28] But we do do some decommissioning work in the oil and gas sector.

[00:08:31] Obviously focusing on the energy transition, we need to clean up anything that's been put out,

[00:08:36] out to sea already and get that removed.

[00:08:38] So decommissioning is quite a key point.

[00:08:40] Now, we have technology that overarches all of those different stages,

[00:08:46] some of them more applicable to others.

[00:08:47] But if we look at the operations and maintenance base again, we have a really strong deployment of autonomy.

[00:08:54] So a subsea robot typically takes nine people to operate and interpret the data that's being seen there.

[00:09:01] If you imagine a seven story vessel about 70 meters long deploying a two meter cube with some propellers on it to fly around.

[00:09:08] And ultimately, in the traditional sense, just to deploy a camera, which some person speaks over relatively low tech.

[00:09:15] What we do is we think, well, first of all, you need to perceive the environment.

[00:09:19] So we've developed SubSlam 3D reconstruction tool, but also subsea compute, which allows us to run this agile AI, but also run autonomy.

[00:09:29] So rather than use nine people to deploy a robot, we take that same similar platform, deploy SubSlam on it and our autonomy platform,

[00:09:38] and allow a robot to autonomously inspect an asset by itself without any human intervention.

[00:09:44] Clearly, that's then massively speeding up the time it takes to do these inspections

[00:09:50] and ultimately how long it takes us to deliver our projects.

[00:09:54] And you don't have to mention any names here, but who are your customers?

[00:09:58] What kind of industries are they in?

[00:10:00] And also what kind of regions are we?

[00:10:01] We're talking in Lisbon in Portugal here.

[00:10:04] But looking further afield, we have a lot of US listeners as well.

[00:10:07] What markets do you serve and who are your customers?

[00:10:10] Yeah.

[00:10:10] So our customers are probably on the top of your electricity bill that comes through the door.

[00:10:15] We're a UK based company.

[00:10:17] So predominantly our work is done in the UK, but we've done a lot of work in Europe as well.

[00:10:21] And that's certainly a growing scene and a growing industry.

[00:10:24] In fact, European offshore wind growth is about three, three and a half times that of the UK.

[00:10:30] So clearly a large market that we're working into.

[00:10:33] We've worked in 19 countries globally, including the US, including countries in the Asia Pacific region.

[00:10:40] So essentially we are following the growth of wind as it grows globally.

[00:10:44] And how much an impact has technology had on the industry?

[00:10:48] If we go back, I don't know what, three, four, five years to what it is now.

[00:10:52] Is technology having that much of an impact?

[00:10:55] Technology is certainly having an impact in the offshore wind space.

[00:10:58] When we started eight years ago, the scopes were very basic.

[00:11:02] So it would be simply we want a video of this asset and this component.

[00:11:06] We're seeing more and more technology that we've developed like subslam these 3D capabilities being built into the scopes of work.

[00:11:13] So the customers are now actually demanding that we come out and do these 3D surveys.

[00:11:18] Also, the way that data is delivered is is becoming massively changed.

[00:11:23] It used to be 100 page PDF reports in literary pitches, text tables that someone would have to go through and review manually.

[00:11:31] And we've got Vantage now, which is a data platform that our customers log into, are able to go asset by asset, look at all of the different images, reports, 3D models into a separate space.

[00:11:43] So tech is hugely changing the way that our customers are thinking about the wind farm and the next steps.

[00:11:49] And if we talk about AI and any big tech project now because of the economic uncertainty, it's always ROI on the project.

[00:11:56] Okay, we'll probably start with wants to return on my investment here.

[00:11:59] Is there anything you can share around that for anyone that may be listening, thinking about beam and then looking up the kind of work that you do and the kind of value that you can offer them?

[00:12:08] So beyond the deliverables that our customers are asking for generally as part of our initial campaigns, one of the real benefits that we bring is quantitative data that comes from these 3D reconstructions.

[00:12:20] So I keep backing on about video being the main deliverable, but really for the last five or six decades, it has been the only thing people have been getting.

[00:12:28] You can imagine video from 30, 40 years ago is particularly grainy.

[00:12:32] If you want to look at something from a certain angle, you haven't necessarily got it.

[00:12:36] What you've got is what you've got.

[00:12:38] You can't do much more with that.

[00:12:39] The 3D data that we get from SubSlam is really future-proofing future development.

[00:12:45] So you might not need a 3D reconstruction of the asset this year because everything's all good.

[00:12:50] But when we come back in three years' time, we do the same asset again.

[00:12:54] We can do a comparison to submillimetric precision and see what changes might be.

[00:12:59] And from that, you might say, oh, well, actually, this thing that we thought was okay actually isn't okay.

[00:13:04] It's definitely changed.

[00:13:05] And then we have all of the data that we've had from these sites historically.

[00:13:10] We do that comparison over multiple assets.

[00:13:12] So you can spot trends in these different assets over time.

[00:13:16] So it's not just about the value we're delivering for the initial project.

[00:13:20] It's also about how we can leverage that data that we've collected on historical projects and apply that to current trends.

[00:13:26] And if we were to zoom out for a moment, what's your grand vision for me?

[00:13:30] What do you want to achieve at world superpower?

[00:13:33] Clearly.

[00:13:34] But no, I mean, we're really keen to support the growth of the offshore wind industry.

[00:13:41] There's clearly going to come a tipping point where the price of oil and gas,

[00:13:45] because it's becoming less and less desirable as an energy source,

[00:13:49] I anticipate that the cost of oil and gas will continue to decline.

[00:13:53] The cost to generate it are relatively fixed, and there's certain states and companies that are able to essentially set the price for oil as they see fit.

[00:14:01] So my personal ambition, and I hope that applies then to Beam, is to allow wind to continue to compete on the global stage for energy.

[00:14:10] Clearly with wind, electric, you're generating electricity at the source.

[00:14:15] You're unable to transmit it.

[00:14:17] Unless we somehow develop massive ships with big batteries and start shipping around the planet,

[00:14:22] we need wind to be viable and cost-effective on a local scale.

[00:14:27] So, for example, the operations and maintenance phase of a wind farm is around 25 years long.

[00:14:34] It's kind of this mid-period of, brilliant, everything's up and running, the turbines are spinning,

[00:14:38] the electricity's being generated.

[00:14:40] The operators are trying to save as much money as possible to make sure that they can be profitable.

[00:14:44] Ultimately, they're a business at the end of the day as well.

[00:14:47] If we look at the operations and maintenance phase of a wind farm, it's about 20 years long.

[00:14:51] It's when everything's up and running, the turbines are spinning, we're getting electricity out of the wind farm.

[00:14:56] Ultimately, the wind farm is a business, and they need to generate revenue.

[00:14:59] So through that 20-year period, they're trying to spend as little money as possible.

[00:15:03] We have a product coming out very, very shortly that really revolutionizes this area of wind

[00:15:09] to make it much, much cheaper, much more cost-effective to do the operations and maintenance type inspections

[00:15:14] that are needed through that period.

[00:15:16] And that's really helping offshore wind to remain competitive in a potential market

[00:15:21] where oil and gas prices continue to drop.

[00:15:23] Exciting times ahead.

[00:15:25] And obviously, we're here at Web Summit, Elizabeth, and I've got to ask,

[00:15:28] what brought you to Web Summit?

[00:15:30] Is it keynotes?

[00:15:32] You have an exhibition stand, press releases, all of the above.

[00:15:35] What brings you here?

[00:15:36] So I think Web Summit's tagline is where tech comes to be born.

[00:15:40] I think really that's actually what we're here for.

[00:15:42] We have a really interesting press conference that's happening tomorrow on the Wednesday, Wednesday morning.

[00:15:48] So I think by the time this comes out, that'll all be live, and we can talk about it in a bit more detail.

[00:15:52] But we've also got a stand, and actually this is the first morning, and it's been incredibly busy.

[00:15:57] It's really great to speak to a whole host of different people coming from an incredibly wide number of regions.

[00:16:06] But also sectors.

[00:16:08] I mean, I've talked about solutions to problems that your audience maybe didn't know existed.

[00:16:13] I've already talked about problems that I didn't know existed already.

[00:16:17] So it's a real melting pot, and it's been really interesting to get stuck in with that.

[00:16:22] Obviously, only into the morning, so I've got another two and a half days to really get stuck in.

[00:16:26] But there's a huge amount of talks, which I'm really interested to get to see.

[00:16:30] And assuming this does go live Wednesday or the Thursday, and somebody from Web Summit could be listening to this,

[00:16:35] where is your exhibition, so where can people find you there?

[00:16:38] We're right by the front entrance.

[00:16:39] If you head in through the main front door, I think you go two or three rows back, we're there on the right.

[00:16:44] And obviously another big theme of Web Summit is the networking side of it,

[00:16:47] and just being able to bounce ideas off people from maybe even outside of industries that we're in.

[00:16:52] And as you said a moment ago, a big melting pot.

[00:16:54] So what are you going to be thinking about on that plane ride home from all the conversation you've been having

[00:16:59] and everything that you've seen on the show floor here?

[00:17:01] Yeah, that's really coming together for me at the moment.

[00:17:04] Still having conversations now, so it's still coming together.

[00:17:07] But it's clear that AI is really here to stay.

[00:17:11] I think it's been a real buzzword in the last 18 to 24 months.

[00:17:15] I think what we're developing and what we're starting to see from elsewhere

[00:17:19] is actual real-world proper application of this kind of technology

[00:17:24] and this kind of ability to augment what a person is able to do with their time and speed up work.

[00:17:31] So thoughts around we have these AI pipelines already, but what else can we be doing with those?

[00:17:37] What are there other potential segments that we might not be applying ourselves to?

[00:17:42] And I'm always on the thought train of, okay, what is this person, what is it that they do?

[00:17:48] And how can Beam help that?

[00:17:50] You know, we might not necessarily be in the offshore wind space,

[00:17:53] but is it a potential industry that we could look to move into?

[00:17:56] And really just trying to get some insight from them as to how could Beam help with their industry.

[00:18:02] Fantastic.

[00:18:03] And outside of Web Summit, people at home listening anywhere in the world,

[00:18:07] they want to find out more information about Beam.

[00:18:09] Maybe even have a question for you or your team.

[00:18:11] Well, do you like to point everyone?

[00:18:13] Yeah, so our website is beam.global.

[00:18:15] Relatively easy to find.

[00:18:16] It's a bright pink homepage, so you should find that relatively easily.

[00:18:19] I'm on LinkedIn, Joe Titballer on LinkedIn, and any of our social media channels.

[00:18:25] Well, I know how incredibly busy you are here at Web Summit,

[00:18:28] and I've learned so much from you today.

[00:18:31] And I'll add links to everything so people can find you nice and easily.

[00:18:34] But more than anything, thanks for stopping by today.

[00:18:36] Yeah, thank you very much for your time.

[00:18:37] Bye for now.

[00:18:38] Speaking with Joe today, I think has given us all a glimpse into the future of renewable energy

[00:18:45] and helped us understand how Beam's pioneering use of AI and autonomy

[00:18:49] is already reshaping the offshore wind industry.

[00:18:53] And that blend of cutting-edge technology with a clear mission to address some of the industry's

[00:19:00] biggest challenges highlights how deep tech innovations is already paving the way for a more sustainable future.

[00:19:08] And as Beam continues to push the boundaries with their autonomous robotic solutions,

[00:19:14] the bigger question is, will AI and automation become the backbone of offshore energy,

[00:19:23] ensuring we meet the global climate goals?

[00:19:27] Love to hear your thoughts on this one and this exciting development.

[00:19:31] It's the future of wind energy in the hands of AI.

[00:19:35] Please share your perspectives and stay tuned for more inspiring conversations from Web Summit this week.

[00:19:41] So email me, techplogwriteroutlook.com, LinkedIn X, Instagram just at Neil C. Hughes.

[00:19:48] Let's keep this conversation going.

[00:19:50] But as for me, back to the show floor I go to find another great guest.

[00:19:54] Speak with you all tomorrow.