1 00:00:05,400 --> 00:00:08,119 Speaker 1: Welcome to the Business of Tech for Business desk empowered 2 00:00:08,119 --> 00:00:11,520 Speaker 1: by two Degrees. I'm Peter Griffin and today I'm catching 3 00:00:11,600 --> 00:00:14,880 Speaker 1: up with Jack Curtis, one of the co founders of 4 00:00:14,960 --> 00:00:20,479 Speaker 1: Australia's newest unicorn startup, Nearer, which now has evaluation of 5 00:00:20,600 --> 00:00:23,920 Speaker 1: over one billion dollars Australian. It's been around for less 6 00:00:23,960 --> 00:00:29,200 Speaker 1: than six years. Nearer uses artificial intelligence to reinvent global 7 00:00:29,320 --> 00:00:32,519 Speaker 1: power grids. It as customers all over the world, including 8 00:00:32,520 --> 00:00:36,800 Speaker 1: here in New Zealand, and raised about ninety million Australian 9 00:00:37,040 --> 00:00:40,479 Speaker 1: last month in its Series D round, led by US 10 00:00:40,479 --> 00:00:46,199 Speaker 1: investor TCV. Now they've invested in the likes of Facebook, Netflix, 11 00:00:46,240 --> 00:00:50,159 Speaker 1: Spotify and our own zero. Why is Nearer of interest 12 00:00:50,159 --> 00:00:52,839 Speaker 1: to us? Well, we've talked a lot in this country 13 00:00:52,920 --> 00:00:57,080 Speaker 1: over the last decade about digital twins. These are digital 14 00:00:57,120 --> 00:00:59,880 Speaker 1: representations of the physical world that can be used for 15 00:01:00,000 --> 00:01:04,360 Speaker 1: so much more accurate and efficient planning and operation of 16 00:01:04,600 --> 00:01:08,679 Speaker 1: assets and infrastructure. But it's been a slow revolution to 17 00:01:08,720 --> 00:01:12,200 Speaker 1: take hold. The energy sector is a prime candidate for 18 00:01:12,280 --> 00:01:17,320 Speaker 1: using physics based modeling to run infrastructure. Imagine this sort 19 00:01:17,319 --> 00:01:21,240 Speaker 1: of scenario. You're using three D modeling and imagery to 20 00:01:21,319 --> 00:01:25,360 Speaker 1: map every inch of your electricity network, including every tree 21 00:01:25,560 --> 00:01:30,880 Speaker 1: in proximity to your electricity lines and transformers. Then, based 22 00:01:30,920 --> 00:01:33,720 Speaker 1: on the known characteristics of those trees, you can model 23 00:01:33,720 --> 00:01:36,880 Speaker 1: what happens if a southerly blast produces gusts of one 24 00:01:36,959 --> 00:01:39,600 Speaker 1: hundred and twenty kilometers per hour. Are the trees going 25 00:01:39,640 --> 00:01:41,560 Speaker 1: to go down? Are they going to take out your 26 00:01:41,600 --> 00:01:45,039 Speaker 1: lines in the process. What happens if a one in 27 00:01:45,120 --> 00:01:48,800 Speaker 1: twenty year flood occurs. How will the floodwaters affect your 28 00:01:48,800 --> 00:01:53,840 Speaker 1: substations and transformers. How quickly will the water recede, What 29 00:01:54,000 --> 00:01:56,160 Speaker 1: impact will there be on the grid. This is an 30 00:01:56,200 --> 00:01:58,760 Speaker 1: interesting scenario if you turn on a new data center 31 00:01:58,800 --> 00:02:02,360 Speaker 1: that is running almost six exclusively AI workloads that have 32 00:02:02,640 --> 00:02:06,360 Speaker 1: very high power requirements. These are the sorts of questions 33 00:02:06,440 --> 00:02:10,960 Speaker 1: you can answer in detail with physics based digital twins. 34 00:02:11,120 --> 00:02:16,360 Speaker 1: The same goes for transport, for our underground infrastructure, evenports, hospitals, 35 00:02:16,360 --> 00:02:20,760 Speaker 1: and airports. We've had GIS screens and models in place 36 00:02:20,840 --> 00:02:23,560 Speaker 1: for a long time, but there's sort of increasingly blunt 37 00:02:23,560 --> 00:02:29,880 Speaker 1: tools when you're dealing with extreme weather floods, earthquakes, even bushfires, 38 00:02:30,200 --> 00:02:33,239 Speaker 1: and spiky new demand from the likes of data centers 39 00:02:33,400 --> 00:02:38,480 Speaker 1: and the electrification off transport and our industry as well. 40 00:02:38,680 --> 00:02:41,400 Speaker 1: So what if you could build a living, breathing, physics 41 00:02:41,400 --> 00:02:45,400 Speaker 1: based model of the entire grid, every pole line and 42 00:02:45,480 --> 00:02:48,400 Speaker 1: transformer and watch how it actually behaves in a storm 43 00:02:48,560 --> 00:02:50,600 Speaker 1: or when you plug in and you wind farm or 44 00:02:50,720 --> 00:02:55,079 Speaker 1: aluminum smelter sized data center. That's what Nearer specializes in. 45 00:02:55,160 --> 00:02:58,679 Speaker 1: It's already modeled more than fifteen million assets covering over 46 00:02:58,720 --> 00:03:03,000 Speaker 1: three million kilometers of infrastructure across four continents. In Australia, 47 00:03:03,000 --> 00:03:06,919 Speaker 1: where it's based, roughly ninety percent off network utilities now 48 00:03:07,000 --> 00:03:10,920 Speaker 1: use its platform, making physics enabled digital twins sort of 49 00:03:10,919 --> 00:03:14,359 Speaker 1: like the fact those standard really for grid planning and 50 00:03:14,639 --> 00:03:17,919 Speaker 1: risk management. Here in New Zealand we're a bit further 51 00:03:17,960 --> 00:03:20,320 Speaker 1: behind on this kind of tech, but with a lot 52 00:03:20,320 --> 00:03:24,080 Speaker 1: of renewables in the pipeline, big data centers coming online, 53 00:03:24,120 --> 00:03:26,679 Speaker 1: and one in one hundred storms that now seem to 54 00:03:26,800 --> 00:03:30,280 Speaker 1: arrive every year or two, physics based digital twins are 55 00:03:30,280 --> 00:03:33,120 Speaker 1: definitely something we should be seriously looking at, not just 56 00:03:33,440 --> 00:03:37,160 Speaker 1: in electricity as well. So. Jack Curtis is a co 57 00:03:37,240 --> 00:03:40,680 Speaker 1: founder and chief commercial officer at Nearer. Jack spent over 58 00:03:40,720 --> 00:03:43,800 Speaker 1: a decade in the solar industry in the US before 59 00:03:43,840 --> 00:03:48,040 Speaker 1: helping build what's now a category defining digital twin platform 60 00:03:48,120 --> 00:03:51,680 Speaker 1: for critical infrastructure. We'll talk about how Nearer started as 61 00:03:51,680 --> 00:03:55,360 Speaker 1: a sort of weekend project for a powerline design tool, 62 00:03:56,000 --> 00:03:59,640 Speaker 1: how AI is being used to build more accurate models faster, 63 00:04:00,160 --> 00:04:03,360 Speaker 1: and what lessons New Zealand can take from utilities overseas 64 00:04:03,360 --> 00:04:07,680 Speaker 1: that are already digitizing their way through and infrastructure and 65 00:04:07,760 --> 00:04:25,039 Speaker 1: climate crisis. Let's get into it, Jack Curtis, Welcome to 66 00:04:25,040 --> 00:04:28,320 Speaker 1: the business of tech. Thanks for having me better well, 67 00:04:28,320 --> 00:04:32,159 Speaker 1: congratulations Series d rays nineteen million dollars. What a great 68 00:04:32,200 --> 00:04:33,080 Speaker 1: way to start the year. 69 00:04:33,920 --> 00:04:36,680 Speaker 2: Yeah, no, it could be worse. But looking forward to 70 00:04:37,400 --> 00:04:40,760 Speaker 2: getting back to the business of putting it, putting it 71 00:04:40,800 --> 00:04:44,080 Speaker 2: into the market, into customers, and into the product. 72 00:04:44,760 --> 00:04:48,599 Speaker 1: Yeah, a remarkable story. You're about well five years, five 73 00:04:48,640 --> 00:04:51,919 Speaker 1: and a half years into this journey with ner. You 74 00:04:51,960 --> 00:04:54,800 Speaker 1: didn't start out sort of as an engineer or anything, right, 75 00:04:54,839 --> 00:04:58,520 Speaker 1: You have a legal background originally exactly so. 76 00:04:58,920 --> 00:05:01,560 Speaker 2: My co founder Dan, who's actually the original founder of 77 00:05:01,600 --> 00:05:04,560 Speaker 2: the business, he has a background in software engineering, and 78 00:05:05,040 --> 00:05:07,760 Speaker 2: it's a quite a funny genesis story. His wife is 79 00:05:07,800 --> 00:05:11,040 Speaker 2: a power line designer and way back when she was 80 00:05:11,120 --> 00:05:15,520 Speaker 2: bemoaning the functionality of existing solutions, and so he kind 81 00:05:15,520 --> 00:05:17,839 Speaker 2: of asked what the problem was and tried to build 82 00:05:17,839 --> 00:05:20,200 Speaker 2: her something better over a weekend, and it turns out 83 00:05:20,200 --> 00:05:24,440 Speaker 2: that it was and then he was essentially I'm glad 84 00:05:24,480 --> 00:05:26,599 Speaker 2: you're happy, honey, and went back to his day job. 85 00:05:27,200 --> 00:05:30,080 Speaker 2: And then oddly it got organic pickup in an industry 86 00:05:30,400 --> 00:05:34,080 Speaker 2: that's not renowned for organic adoption of anything. And so 87 00:05:34,200 --> 00:05:37,640 Speaker 2: that was way back in twenty sixteen, and then it 88 00:05:37,760 --> 00:05:39,680 Speaker 2: kind of got enough traction where he thought it was 89 00:05:39,680 --> 00:05:43,760 Speaker 2: probably worth turning into a company. And then Karm other 90 00:05:43,839 --> 00:05:47,480 Speaker 2: co founder joined, and I joined shortly after that, and 91 00:05:47,520 --> 00:05:49,920 Speaker 2: then it kind of became a business. And that was 92 00:05:49,960 --> 00:05:51,320 Speaker 2: around yeah, five six years ago. 93 00:05:51,360 --> 00:05:54,400 Speaker 1: Now, yeah, prior to that, you did about what twelve 94 00:05:54,480 --> 00:05:58,640 Speaker 1: years at first solo one of the big solar panel 95 00:05:58,760 --> 00:06:01,719 Speaker 1: manufacturers in the US would have been a fascinating time 96 00:06:01,760 --> 00:06:04,719 Speaker 1: given compared to New Zealand. I guess, given the big 97 00:06:04,800 --> 00:06:07,599 Speaker 1: uptake of solar during that decade. 98 00:06:08,040 --> 00:06:11,240 Speaker 2: Yeah, it really was. I joined right at the I guess, 99 00:06:11,320 --> 00:06:16,480 Speaker 2: outset of the renewable energy trajectory from being quite expensive 100 00:06:17,120 --> 00:06:21,080 Speaker 2: now obviously being some of the lowest cost electricity on Earth, 101 00:06:21,120 --> 00:06:23,479 Speaker 2: whether it's solar or wind or whatever it might be, 102 00:06:24,000 --> 00:06:29,120 Speaker 2: and really lived through that transition period from when renewal 103 00:06:29,200 --> 00:06:32,720 Speaker 2: energy was very reliant on subsidies and policy support to 104 00:06:32,760 --> 00:06:35,880 Speaker 2: the point where it became very commercially viable. And so 105 00:06:36,080 --> 00:06:39,560 Speaker 2: through that period at first solar ended up operating in 106 00:06:39,760 --> 00:06:42,280 Speaker 2: close to about thirty different countries around the world and 107 00:06:42,360 --> 00:06:46,000 Speaker 2: really got to see how different markets, different governments thought 108 00:06:46,040 --> 00:06:49,719 Speaker 2: about the energy transition challenge. Some of the things that 109 00:06:50,480 --> 00:06:54,880 Speaker 2: were originally probably a bit overhyped from a constraint point 110 00:06:54,880 --> 00:06:57,240 Speaker 2: of view, and then you know, over time have certainly 111 00:06:57,279 --> 00:07:00,360 Speaker 2: proven to be far more viable and anticipated. I think 112 00:07:00,400 --> 00:07:03,440 Speaker 2: r at the point now where generation went on this 113 00:07:03,520 --> 00:07:07,640 Speaker 2: journey of becoming fit for purpose and sufficiently cost effective, 114 00:07:08,080 --> 00:07:12,200 Speaker 2: and now we're kind of facing other challenges. Relates to realizing, 115 00:07:12,320 --> 00:07:14,920 Speaker 2: you know, a lot of these ambitious energy transition goals. 116 00:07:15,200 --> 00:07:16,840 Speaker 1: Yeah, and we're going to talk about that in the 117 00:07:16,840 --> 00:07:19,360 Speaker 1: context of New Zealand, because boy, we do have some 118 00:07:19,800 --> 00:07:22,480 Speaker 1: challenges but big opportunities and starting from a good base 119 00:07:22,520 --> 00:07:26,560 Speaker 1: where in terms of electricity generated going into the grid, 120 00:07:27,040 --> 00:07:29,920 Speaker 1: you know, eighty percent of it's renewable. So but we 121 00:07:30,000 --> 00:07:32,400 Speaker 1: do have some issues we'll talk about that, but really 122 00:07:32,680 --> 00:07:34,720 Speaker 1: take us back to when you were setting up near 123 00:07:35,880 --> 00:07:38,240 Speaker 1: give us a snapshot of what you know, the digital 124 00:07:38,280 --> 00:07:42,040 Speaker 1: twin world looked like when it came to utilities, which 125 00:07:42,120 --> 00:07:45,200 Speaker 1: was the logical place really to take up that technology. 126 00:07:45,240 --> 00:07:46,800 Speaker 1: I've been talking about it for a long time, but 127 00:07:46,840 --> 00:07:49,840 Speaker 1: it really has been I think in this country slow 128 00:07:48,880 --> 00:07:53,800 Speaker 1: to be adopted, and in other industry is definitely much 129 00:07:53,840 --> 00:07:57,720 Speaker 1: slower transport, for instance. But what did digital twins look 130 00:07:57,840 --> 00:08:01,080 Speaker 1: like in the utility sector Before you start and you 131 00:08:01,160 --> 00:08:05,280 Speaker 1: talk about infrastructure intelligence, give us a snapshot off the 132 00:08:05,440 --> 00:08:09,400 Speaker 1: changes that you thought needed to happen to modeling infrastructure. 133 00:08:09,800 --> 00:08:13,720 Speaker 2: Yeah, it's a great question because the whole digital twin nomenclature, 134 00:08:14,600 --> 00:08:18,920 Speaker 2: it's a very broad church kind of term and almost 135 00:08:19,000 --> 00:08:23,360 Speaker 2: to the point where it's been somewhat bastardized and become 136 00:08:23,400 --> 00:08:26,360 Speaker 2: a buzzword that's lost a bit of meaning. And so 137 00:08:26,440 --> 00:08:29,600 Speaker 2: to your question, when we really entered the space, the 138 00:08:29,720 --> 00:08:35,720 Speaker 2: term digital twin really was more focused on digital modeling 139 00:08:35,760 --> 00:08:38,439 Speaker 2: from a visual point of view, and so you could 140 00:08:38,520 --> 00:08:44,920 Speaker 2: create these very pretty three dimensional models of complex infrastructure 141 00:08:45,000 --> 00:08:48,960 Speaker 2: like electricity grids, and they looked good, but you really 142 00:08:49,000 --> 00:08:54,360 Speaker 2: couldn't prosecute anything of any depth or analysis that would 143 00:08:54,360 --> 00:08:57,640 Speaker 2: give the owners of critical infrastructure the comfort to go 144 00:08:57,920 --> 00:09:01,160 Speaker 2: and execute on that in the physical reund And what 145 00:09:01,200 --> 00:09:04,320 Speaker 2: I mean by that is it's fine to have a 146 00:09:04,360 --> 00:09:08,679 Speaker 2: three D visualization of a car, but if you're operating 147 00:09:08,679 --> 00:09:12,160 Speaker 2: an electricity network that is inherently delivering a critical commodity, 148 00:09:12,679 --> 00:09:15,520 Speaker 2: you can't make a decision just based on a three 149 00:09:15,520 --> 00:09:19,720 Speaker 2: dimensional visual model. And so the first thing that really 150 00:09:20,200 --> 00:09:26,600 Speaker 2: really tried to achieve was communicating the importance of what 151 00:09:26,640 --> 00:09:30,880 Speaker 2: we call a physics digital twin, which is it's not 152 00:09:31,000 --> 00:09:33,680 Speaker 2: just a three dimensional model which is the output, but 153 00:09:33,760 --> 00:09:36,600 Speaker 2: it's actually a behavior model. And what I mean by 154 00:09:36,640 --> 00:09:41,599 Speaker 2: that is when the assets are modeled visually in our platform, 155 00:09:41,760 --> 00:09:44,120 Speaker 2: they are also been infused with all the physics and 156 00:09:44,200 --> 00:09:48,520 Speaker 2: engineering characteristics of those assets. So it's structural integrity how 157 00:09:48,520 --> 00:09:53,600 Speaker 2: it reacts to external events like wind and storms and 158 00:09:53,640 --> 00:09:58,120 Speaker 2: other externalities. And so essentially, when you simulate something in 159 00:09:58,160 --> 00:10:02,160 Speaker 2: a physics digital twin, you're actually getting a very high 160 00:10:02,559 --> 00:10:06,520 Speaker 2: accuracy around how it's going to behave if you actually 161 00:10:06,559 --> 00:10:08,719 Speaker 2: were going to go and invest in an asset in 162 00:10:08,760 --> 00:10:11,440 Speaker 2: the field, or if an extreme weather event happened in 163 00:10:11,440 --> 00:10:14,840 Speaker 2: the field, and so that was really the distinction we 164 00:10:14,840 --> 00:10:17,840 Speaker 2: were looking to drive. Now. The challenge was that when 165 00:10:17,840 --> 00:10:21,559 Speaker 2: we entered the market, there'd been a lot of attempted 166 00:10:22,320 --> 00:10:26,760 Speaker 2: efforts to roll out digital twins to the electricity network space, 167 00:10:26,800 --> 00:10:29,840 Speaker 2: to other critical infrastructure, and because they hadn't really achieved 168 00:10:29,840 --> 00:10:32,360 Speaker 2: what they promised, it had left a bit of a 169 00:10:32,720 --> 00:10:35,440 Speaker 2: I guess, a skepticism, and so we had to overcome 170 00:10:35,480 --> 00:10:38,320 Speaker 2: that skepticism. It was at the time when the space 171 00:10:38,440 --> 00:10:40,360 Speaker 2: was becoming very popular, so there was a lot of 172 00:10:40,440 --> 00:10:43,680 Speaker 2: venture backed companies like US around and so it became 173 00:10:44,120 --> 00:10:47,199 Speaker 2: very noisy, very busy, and for the next kind of 174 00:10:47,240 --> 00:10:50,680 Speaker 2: two to three years after that initial commercial deployment period, 175 00:10:51,280 --> 00:10:53,640 Speaker 2: we really had to navigate a lot of signal to 176 00:10:53,679 --> 00:10:57,440 Speaker 2: noise ratio and that's now started to flush out. But 177 00:10:57,480 --> 00:11:00,320 Speaker 2: I think to the original question, there's now a much 178 00:11:00,320 --> 00:11:04,480 Speaker 2: more precise understanding of the different categories of digital twin. 179 00:11:04,640 --> 00:11:07,560 Speaker 2: What's a visual twin, what's a behavior twin, what's an 180 00:11:07,600 --> 00:11:10,959 Speaker 2: operational twin? And I think the industry has gone on 181 00:11:11,000 --> 00:11:14,000 Speaker 2: a bit of a learning journey to understand those distinctions. 182 00:11:14,559 --> 00:11:19,320 Speaker 1: Yeah, and it's been a lot of GIS material you know, 183 00:11:19,400 --> 00:11:22,400 Speaker 1: visual material available have been into these offices and some 184 00:11:22,440 --> 00:11:24,679 Speaker 1: of these utilities and seen it all looks free fancy. 185 00:11:24,720 --> 00:11:28,600 Speaker 1: But that next step to physics enabled digital twins as 186 00:11:28,720 --> 00:11:32,760 Speaker 1: the potential game changer, and I guess that brings up 187 00:11:32,920 --> 00:11:36,640 Speaker 1: artificial intelligence and machine learning. How are you employing those 188 00:11:36,679 --> 00:11:39,720 Speaker 1: technologies to build that sort of physics enabled digital twin. 189 00:11:40,360 --> 00:11:42,240 Speaker 2: Yeah, so there's a couple of things that we always 190 00:11:42,280 --> 00:11:45,560 Speaker 2: like to stress. The first thing is that the idea 191 00:11:45,920 --> 00:11:50,480 Speaker 2: of utilities using a black box AI model is not 192 00:11:50,520 --> 00:11:54,320 Speaker 2: a great idea. We have always sought to build a 193 00:11:54,360 --> 00:11:59,000 Speaker 2: product where our customers, the owners of electricity grids, could 194 00:11:59,040 --> 00:12:02,040 Speaker 2: always determ and exactly where the analysis had come from, 195 00:12:02,320 --> 00:12:05,160 Speaker 2: so it wasn't just pump data in some AI black 196 00:12:05,200 --> 00:12:08,520 Speaker 2: box gives you a recommendation that's obviously not going to 197 00:12:08,800 --> 00:12:13,040 Speaker 2: pass muster with folks that are operating very complex infrastructure. 198 00:12:13,600 --> 00:12:16,840 Speaker 2: And so the first answer the question is that at 199 00:12:16,880 --> 00:12:21,200 Speaker 2: the end of the day, our platform still primarily function 200 00:12:21,400 --> 00:12:23,800 Speaker 2: the back of being a physics engineering model. Now, what 201 00:12:23,840 --> 00:12:28,000 Speaker 2: we do is that we leverage AI and mL to 202 00:12:28,040 --> 00:12:31,040 Speaker 2: build the models faster, So we use a whole spectrum 203 00:12:31,040 --> 00:12:33,640 Speaker 2: of data, we use lighter, we take in JS, we 204 00:12:33,760 --> 00:12:37,920 Speaker 2: take in satellite imagery, high res imagery, but our ability 205 00:12:37,960 --> 00:12:41,480 Speaker 2: to build those models to be extremely accurate is now 206 00:12:41,640 --> 00:12:45,160 Speaker 2: very enabled by being able to train machine learning models 207 00:12:45,240 --> 00:12:47,520 Speaker 2: on all. Right, we're going to give you a bunch 208 00:12:47,520 --> 00:12:51,000 Speaker 2: of dots that look like nothing, but you've now been 209 00:12:51,040 --> 00:12:54,000 Speaker 2: trained to identify that that's a power pole instead of 210 00:12:54,000 --> 00:12:57,920 Speaker 2: a tree, and so it's helped us build very sophisticated 211 00:12:57,960 --> 00:13:01,520 Speaker 2: models with much greater accuracy, much more efficiently. We also 212 00:13:01,600 --> 00:13:06,520 Speaker 2: now use them to do what we call backcasting forecasting 213 00:13:06,800 --> 00:13:09,200 Speaker 2: in the context of extreme weather. So we now do 214 00:13:09,280 --> 00:13:12,360 Speaker 2: a lot of work across Texas in the context of 215 00:13:13,000 --> 00:13:15,520 Speaker 2: looking at all the extreme weather events that have occurred, 216 00:13:16,080 --> 00:13:19,840 Speaker 2: the damage that was created on infrastructure because of those events, 217 00:13:20,400 --> 00:13:25,560 Speaker 2: and then being able to deterministically tell networks that when 218 00:13:25,600 --> 00:13:29,079 Speaker 2: these storms neverbly occur in the future, here's where the 219 00:13:29,160 --> 00:13:33,000 Speaker 2: damage is going to manifest. And it's not a probabilistic analysis. 220 00:13:33,000 --> 00:13:36,120 Speaker 2: It's actually saying, because it's a behavior model, if this 221 00:13:36,280 --> 00:13:39,719 Speaker 2: storm of this windspeed happens again, here's the assets that 222 00:13:39,760 --> 00:13:42,240 Speaker 2: will fail first. And so what that means from a 223 00:13:42,280 --> 00:13:44,920 Speaker 2: capital allocation point of view is that the utilities can 224 00:13:44,960 --> 00:13:48,080 Speaker 2: say all right, where's the first ten to twenty percent 225 00:13:48,280 --> 00:13:52,000 Speaker 2: of most impactful damage? Is that next to hospitals, next 226 00:13:52,000 --> 00:13:54,679 Speaker 2: to schools? Is it where I have a high concentration 227 00:13:54,760 --> 00:13:57,920 Speaker 2: of customers and they can with far greater precision go 228 00:13:58,040 --> 00:14:01,680 Speaker 2: and preemptively invest in networks to essentially say, I'm going 229 00:14:01,760 --> 00:14:04,360 Speaker 2: to buy down that first twenty percent of the greatest 230 00:14:04,440 --> 00:14:07,040 Speaker 2: risk to my customer base. So we use a lot 231 00:14:07,080 --> 00:14:10,720 Speaker 2: of artificial intelligence in that context. And now with the 232 00:14:10,760 --> 00:14:13,920 Speaker 2: advent you know more recently of what's happening with Anthropic 233 00:14:13,960 --> 00:14:17,360 Speaker 2: and some of the functionality developed you know in that 234 00:14:17,480 --> 00:14:20,240 Speaker 2: latest release, we're now looking at some of the internal 235 00:14:20,280 --> 00:14:25,440 Speaker 2: working ways that we can code faster, deliver models faster 236 00:14:25,480 --> 00:14:30,640 Speaker 2: to customers. And so I think in very complex domains 237 00:14:30,760 --> 00:14:35,520 Speaker 2: like critical infrastructure, particularly electricity networks, the idea that artificial 238 00:14:35,560 --> 00:14:38,240 Speaker 2: intelligence can just be rolled out and solve a bunch 239 00:14:38,280 --> 00:14:42,760 Speaker 2: of problems I don't think is quite accurate. I think 240 00:14:42,800 --> 00:14:46,320 Speaker 2: it's making sure you use it for where it can 241 00:14:46,360 --> 00:14:50,080 Speaker 2: make an impact, whether it's driving greater efficiency, greater accuracy, 242 00:14:51,240 --> 00:14:54,280 Speaker 2: better fidelity, and then starting to look like components of 243 00:14:54,320 --> 00:14:59,880 Speaker 2: the workflows where it can introduce improvements, but with that 244 00:15:00,280 --> 00:15:03,200 Speaker 2: introducing risk I think that's one of the biggest challenges 245 00:15:03,240 --> 00:15:07,400 Speaker 2: that our customers navigate quite rightfully, is everyone must be 246 00:15:07,480 --> 00:15:10,920 Speaker 2: using AI, but you can't be using it to the 247 00:15:10,960 --> 00:15:14,720 Speaker 2: point when you're operating critical infrastructure that you lose sight 248 00:15:15,120 --> 00:15:18,840 Speaker 2: of the human input and the human decision because ultimately 249 00:15:18,920 --> 00:15:22,000 Speaker 2: they still understand their networks better than a software platform does. 250 00:15:22,640 --> 00:15:27,480 Speaker 1: Yeah, you brought up extreme weather. They're particularly relevant in 251 00:15:27,520 --> 00:15:30,160 Speaker 1: New Zealand at the moment as we record this. Just 252 00:15:30,440 --> 00:15:33,200 Speaker 1: in the last week there's been a massive storm went 253 00:15:33,240 --> 00:15:38,000 Speaker 1: through New Zealand, took out about thirty thousands connections in 254 00:15:38,040 --> 00:15:41,320 Speaker 1: the North Island. One of your customers, Power Co, was 255 00:15:41,360 --> 00:15:44,040 Speaker 1: affected by that did quite well to get things back 256 00:15:44,080 --> 00:15:47,240 Speaker 1: up and running reasonably quickly. But I guess yeah, things 257 00:15:47,280 --> 00:15:52,880 Speaker 1: like vegetation management so trees aren't falling on lines flat 258 00:15:53,120 --> 00:15:57,560 Speaker 1: and fire modeling. Fire, particularly in Australia, has been a 259 00:15:57,640 --> 00:16:01,960 Speaker 1: major issue for utilities. But when you think about what's 260 00:16:02,000 --> 00:16:07,120 Speaker 1: coming in the years and decades ahead, with more extreme weather, 261 00:16:07,280 --> 00:16:12,800 Speaker 1: more frequent extreme weather events, are the utilities actually listening 262 00:16:12,920 --> 00:16:16,680 Speaker 1: when they see this model modeling Because it's so expensive 263 00:16:16,720 --> 00:16:19,680 Speaker 1: and to think long term and proactively, it is quite 264 00:16:20,680 --> 00:16:23,400 Speaker 1: a sort of a mind shift for big businesses that 265 00:16:23,400 --> 00:16:25,360 Speaker 1: are having to spend one hundreds of millions of dollars 266 00:16:25,360 --> 00:16:29,560 Speaker 1: of investment is what they're seeing with Nearer actually encouraging 267 00:16:29,600 --> 00:16:31,920 Speaker 1: them to think more long term about their assets. 268 00:16:32,400 --> 00:16:34,080 Speaker 2: Yeah, it's a really a point, and I think part 269 00:16:34,120 --> 00:16:37,640 Speaker 2: of the challenge is that a lot of these issues 270 00:16:37,680 --> 00:16:41,360 Speaker 2: have been handled or managed in a very analog way, 271 00:16:42,240 --> 00:16:45,000 Speaker 2: very manual inspection way, where it's sending people out to 272 00:16:45,080 --> 00:16:50,320 Speaker 2: inspect risk, identify risk, determine risk, report back, manually execute 273 00:16:50,320 --> 00:16:56,240 Speaker 2: on that, and that almost is becoming it's become a 274 00:16:56,240 --> 00:16:58,080 Speaker 2: bit of an existential crisis for a lot of network 275 00:16:58,160 --> 00:17:02,520 Speaker 2: utilities in California. We work with most of California. Now 276 00:17:03,120 --> 00:17:05,880 Speaker 2: it's just not acceptable to be doing that. And so 277 00:17:06,040 --> 00:17:11,000 Speaker 2: now those entire programs around vegetation management, risk identification, telling 278 00:17:11,119 --> 00:17:14,320 Speaker 2: networks what to cut, when to cut, where their assets are, 279 00:17:14,680 --> 00:17:19,200 Speaker 2: that's all digitized and Nearer. In Texas, it's the same 280 00:17:19,240 --> 00:17:21,920 Speaker 2: with extreme weather. So you know that example I was 281 00:17:21,960 --> 00:17:26,120 Speaker 2: giving about backcasting and forecasting, we now work with pretty 282 00:17:26,200 --> 00:17:29,880 Speaker 2: much every Texas utility because one of our biggest customers 283 00:17:29,880 --> 00:17:33,520 Speaker 2: there they were still doing vegetation management manually and then 284 00:17:33,560 --> 00:17:36,720 Speaker 2: a storm came through all the vegetation fell on the lines, 285 00:17:36,920 --> 00:17:40,399 Speaker 2: powers out for three days, twenty five people tragically died. 286 00:17:40,640 --> 00:17:44,560 Speaker 2: They spent four weeks in front of House Senate inquiries 287 00:17:45,080 --> 00:17:47,879 Speaker 2: and now it's just not acceptable. And so now that 288 00:17:48,080 --> 00:17:51,280 Speaker 2: entire workflow is digitized in nearer and so I think 289 00:17:51,359 --> 00:17:54,240 Speaker 2: the challenge a little bit is that it will soon 290 00:17:54,320 --> 00:17:57,920 Speaker 2: reach a point where if you are a network utility, 291 00:17:58,480 --> 00:18:02,360 Speaker 2: and it is entirely possible to dramatically buy down your 292 00:18:02,440 --> 00:18:07,439 Speaker 2: risk by digitizing these workflows, particularly around extreme weather and 293 00:18:07,520 --> 00:18:12,080 Speaker 2: external weather events, and essentially not only be able to 294 00:18:12,080 --> 00:18:16,200 Speaker 2: buy down the risk, but reduce the spend profile because 295 00:18:16,240 --> 00:18:18,320 Speaker 2: you end up spending so much more on the back 296 00:18:18,440 --> 00:18:21,719 Speaker 2: end rectifying it than you do on the front end 297 00:18:21,800 --> 00:18:25,040 Speaker 2: by identifying where the most likely kind of areas of 298 00:18:25,119 --> 00:18:28,159 Speaker 2: risk are. And so I think the challenge now is 299 00:18:28,200 --> 00:18:31,080 Speaker 2: that we're almost at this tipping point that it's not 300 00:18:31,960 --> 00:18:35,679 Speaker 2: why nearer, it's almost like why not? And if you're 301 00:18:35,720 --> 00:18:41,919 Speaker 2: still managing these issues manually and not sufficiently investing in 302 00:18:42,000 --> 00:18:44,639 Speaker 2: buying down that risk, buying down that cost to consumers, 303 00:18:45,359 --> 00:18:47,760 Speaker 2: consumers will be asking, hang on, we actually could have 304 00:18:47,800 --> 00:18:50,600 Speaker 2: identified this risk before it happened to a large degree. 305 00:18:51,080 --> 00:18:53,400 Speaker 2: We could have spent far less on the front end 306 00:18:53,560 --> 00:18:56,800 Speaker 2: mitigating that risk, and now just because we didn't do that, 307 00:18:57,280 --> 00:19:00,639 Speaker 2: we're now passing on more power out of just consumers, 308 00:19:00,960 --> 00:19:04,240 Speaker 2: we're passing on more cost to consumers and potentially even 309 00:19:04,280 --> 00:19:06,679 Speaker 2: more safety risk. And I think it's been a bit 310 00:19:06,720 --> 00:19:09,679 Speaker 2: of an industry journey, but now at the point where 311 00:19:09,720 --> 00:19:15,119 Speaker 2: in Australia ninety percent of utilities is nearer, we're now 312 00:19:15,280 --> 00:19:19,159 Speaker 2: the standard in the most high exposure states in the 313 00:19:19,200 --> 00:19:22,359 Speaker 2: world in Texas and California. I think New Zealand that 314 00:19:22,400 --> 00:19:28,560 Speaker 2: has been genuinely fairly proficient at navigating energy issues historically, 315 00:19:28,680 --> 00:19:30,439 Speaker 2: and we work with a number of the network untilities 316 00:19:30,480 --> 00:19:33,320 Speaker 2: in New Zealand. We yet to see that real kind 317 00:19:33,359 --> 00:19:35,960 Speaker 2: of like wake up point where it's like this is 318 00:19:36,000 --> 00:19:39,440 Speaker 2: now the standard. It's going to be a big issue 319 00:19:39,480 --> 00:19:43,000 Speaker 2: explaining why we're not digitizing these workflows as opposed to 320 00:19:43,080 --> 00:19:46,040 Speaker 2: still relying on very out of date analog manual processes. 321 00:19:46,359 --> 00:19:49,600 Speaker 1: Absolutely well, it's becoming a unfortunately as sort of an 322 00:19:49,640 --> 00:19:52,840 Speaker 1: annual or biannual event now where we get, particularly an 323 00:19:52,880 --> 00:19:56,080 Speaker 1: eastern part of the country, these extreme weather advance, the 324 00:19:56,160 --> 00:20:00,480 Speaker 1: tropical cyclones coming down causing millions of dollars worth of 325 00:20:00,560 --> 00:20:04,440 Speaker 1: damage every time if they can from an evidence base, say, 326 00:20:04,480 --> 00:20:07,520 Speaker 1: if we invest some money now we can reinforce part 327 00:20:07,520 --> 00:20:10,439 Speaker 1: of our network and to make it more resilient for 328 00:20:11,320 --> 00:20:14,560 Speaker 1: fire flowing communities, but also save our investors and save 329 00:20:14,600 --> 00:20:15,560 Speaker 1: our customers money. 330 00:20:16,840 --> 00:20:20,280 Speaker 2: Exactly right, Mart, and I think that that's what needed 331 00:20:20,320 --> 00:20:25,120 Speaker 2: to change because you know, quite quite legitially historically, efforts 332 00:20:25,119 --> 00:20:29,840 Speaker 2: to digitize you know, assets and analysis and the digital 333 00:20:29,880 --> 00:20:33,600 Speaker 2: twin context didn't quite have the level of accuracy or fidelity. 334 00:20:34,040 --> 00:20:36,680 Speaker 2: But now what we're talking about, it's not even probability, 335 00:20:36,920 --> 00:20:40,920 Speaker 2: it's deterministic because we understand all the behaviors of the assets. 336 00:20:41,520 --> 00:20:44,000 Speaker 2: It's almost like the reductive example I give is that 337 00:20:44,640 --> 00:20:47,280 Speaker 2: if you were looking at a powerpole outside your house, 338 00:20:47,520 --> 00:20:51,000 Speaker 2: you know, during a high wind event, and looked at 339 00:20:51,000 --> 00:20:53,640 Speaker 2: that same pole and nearer it would be doing exactly 340 00:20:53,680 --> 00:20:56,320 Speaker 2: the same thing, bending and flexing and then ultimately snapping. 341 00:20:56,960 --> 00:21:01,360 Speaker 2: And so now we could say to you know, network owner, hey, 342 00:21:01,400 --> 00:21:02,920 Speaker 2: this is going to happen to the pole in front 343 00:21:02,920 --> 00:21:06,239 Speaker 2: of Peter's house. Why don't you spend you know, X 344 00:21:06,280 --> 00:21:10,119 Speaker 2: dollars reinforcing that because we know these extreme weather events 345 00:21:10,160 --> 00:21:12,879 Speaker 2: are just happening on a more regular cycle. So spend 346 00:21:12,960 --> 00:21:16,399 Speaker 2: a dollar reinforcing that pole because we can guarantee you 347 00:21:16,440 --> 00:21:19,400 Speaker 2: it's going to snap, instead of spending you know, ten 348 00:21:19,440 --> 00:21:22,280 Speaker 2: dollars replacing the pole when it snaps. And then you're like, 349 00:21:22,320 --> 00:21:24,159 Speaker 2: hang on, why would I not want the network to 350 00:21:24,160 --> 00:21:25,879 Speaker 2: do that? I don't want to power out be I 351 00:21:25,880 --> 00:21:27,719 Speaker 2: don't like the idea as a consumer of paying ten 352 00:21:27,800 --> 00:21:30,800 Speaker 2: dollars through a pole replacement. And now you can actually 353 00:21:30,800 --> 00:21:34,000 Speaker 2: do that, you know, completely in compliance with all regulatory 354 00:21:34,000 --> 00:21:37,160 Speaker 2: engineering standards, you know, to one hundred percent level of avacuracy. 355 00:21:37,720 --> 00:21:40,600 Speaker 2: And so again now it's like not doing this is 356 00:21:40,640 --> 00:21:41,960 Speaker 2: just a hectic force economy. 357 00:21:42,480 --> 00:21:46,680 Speaker 1: Yeah. So there's some really clear quick ones there that 358 00:21:47,040 --> 00:21:50,560 Speaker 1: you know, your customers were obviously benefiting from a New 359 00:21:50,640 --> 00:21:54,240 Speaker 1: Zealand and around the world. The probably more complex issue 360 00:21:54,280 --> 00:21:56,800 Speaker 1: we're grappling with at the moment is, you know, we've 361 00:21:56,800 --> 00:22:00,440 Speaker 1: got this high share of renewables into the grid eighty 362 00:22:00,440 --> 00:22:03,520 Speaker 1: to ninety percent on a good day, but we have 363 00:22:03,640 --> 00:22:06,560 Speaker 1: these recurring dry winters, the dry year problem where we 364 00:22:06,640 --> 00:22:10,080 Speaker 1: have to fall back on fossil fuels. We've got a 365 00:22:10,080 --> 00:22:13,120 Speaker 1: lot of renewables in the pipeline being consented, so we're 366 00:22:13,119 --> 00:22:14,960 Speaker 1: going to have more wind farms, a little bit of solo, 367 00:22:15,000 --> 00:22:18,200 Speaker 1: and maybe even some deep geothermal if it comes off. 368 00:22:18,840 --> 00:22:22,160 Speaker 1: But it's sort of how all of this stuff interacts, 369 00:22:23,200 --> 00:22:25,919 Speaker 1: you know, making sure we're not overbuilding physical assets, that 370 00:22:25,920 --> 00:22:29,040 Speaker 1: we're not duplicating resources. Can you help, Can your physics 371 00:22:29,080 --> 00:22:33,280 Speaker 1: based digital twin help to alleviate this sort of issue 372 00:22:33,280 --> 00:22:35,359 Speaker 1: we have as we try to make that transition to 373 00:22:35,440 --> 00:22:36,320 Speaker 1: more renewables. 374 00:22:36,680 --> 00:22:38,679 Speaker 2: Yeah, so on one of the big themes that we 375 00:22:38,800 --> 00:22:42,560 Speaker 2: focus on. So essentially we have four big thematic themes 376 00:22:43,840 --> 00:22:46,960 Speaker 2: that the platform supports. One is design and constructing new assets. 377 00:22:47,400 --> 00:22:51,000 Speaker 2: The second one is optimizing existing assets, so how to 378 00:22:51,000 --> 00:22:55,400 Speaker 2: not overspan gold plate assets while still making sure they're performing. 379 00:22:55,600 --> 00:22:59,359 Speaker 2: The third one is this extreme weather theme we're talking about, 380 00:22:59,359 --> 00:23:04,240 Speaker 2: and the fourth one is optimizing the generation network mix 381 00:23:04,280 --> 00:23:07,080 Speaker 2: in the context of an energy transition. And so most 382 00:23:07,080 --> 00:23:10,480 Speaker 2: of the time, the problem statement that we're helping navigate 383 00:23:10,760 --> 00:23:15,320 Speaker 2: is these large backlogs of renewable generating assets that are 384 00:23:15,320 --> 00:23:18,320 Speaker 2: ready to connect, but there's not enough network to connect 385 00:23:18,359 --> 00:23:23,040 Speaker 2: them to. And so we focus that in two contexts. 386 00:23:23,400 --> 00:23:27,280 Speaker 2: One accelerating how new transmission projects can be realized through 387 00:23:27,320 --> 00:23:31,040 Speaker 2: the design, permitting and planning construction phase, but actually more 388 00:23:31,040 --> 00:23:36,719 Speaker 2: interestingly identifying existing network and where it is underutilized from 389 00:23:36,760 --> 00:23:40,480 Speaker 2: an assumption point of view. And so quite often what 390 00:23:40,960 --> 00:23:43,760 Speaker 2: is the case is that all networks have to run 391 00:23:43,800 --> 00:23:47,959 Speaker 2: at a certain temperature, above which it creates safety issues. 392 00:23:48,200 --> 00:23:50,440 Speaker 2: And the more electricity you run through an electricity line, 393 00:23:50,440 --> 00:23:53,639 Speaker 2: the hotter it gets. And so because the safety issue 394 00:23:53,680 --> 00:23:57,400 Speaker 2: is a binary, often networks are making conservative assumption about 395 00:23:57,400 --> 00:23:59,520 Speaker 2: how much electricity they can run through the network. And 396 00:23:59,520 --> 00:24:02,840 Speaker 2: so we can do with simulate at every single power line, 397 00:24:02,880 --> 00:24:05,600 Speaker 2: at an individual level and an integrated full network level, 398 00:24:06,040 --> 00:24:09,040 Speaker 2: how much electricity actually can you run through this network 399 00:24:09,320 --> 00:24:12,639 Speaker 2: without creating those safety issues. And so what that enables 400 00:24:12,720 --> 00:24:16,119 Speaker 2: us to do is say, all right, we're now maximizing 401 00:24:16,760 --> 00:24:21,480 Speaker 2: the operational efficiency of New Zealand's electricity network. Now let's 402 00:24:21,520 --> 00:24:25,080 Speaker 2: overlay the generating profile of assets. And so, first of all, 403 00:24:25,200 --> 00:24:27,200 Speaker 2: you need there to be as much of a correlation 404 00:24:27,320 --> 00:24:30,680 Speaker 2: between generation and load so customers that need to or 405 00:24:30,720 --> 00:24:33,440 Speaker 2: want to use it. The second thing is identifying are 406 00:24:33,520 --> 00:24:36,639 Speaker 2: you citing that generation on the most optimal parts of 407 00:24:36,640 --> 00:24:40,360 Speaker 2: the grid where you're not going to create reliability issues, 408 00:24:41,480 --> 00:24:45,440 Speaker 2: you know, more kind of extreme issues around network outage, 409 00:24:45,840 --> 00:24:49,000 Speaker 2: and then to your question as well, you then try 410 00:24:49,040 --> 00:24:52,480 Speaker 2: and optimize how does this all operate as an integrated 411 00:24:52,880 --> 00:24:56,600 Speaker 2: energy mix. And so one of the tyrannies of New 412 00:24:56,680 --> 00:24:59,520 Speaker 2: Zealand's progression around the energy transition, which is admirable, which 413 00:24:59,560 --> 00:25:02,520 Speaker 2: is getting too such a high penetration of renewables because 414 00:25:02,520 --> 00:25:06,640 Speaker 2: it's had all these organic inherent sources within the country. 415 00:25:06,920 --> 00:25:10,240 Speaker 2: Is all right, we sped to eighty ninety percent, which 416 00:25:10,280 --> 00:25:14,320 Speaker 2: is great, But were we able to architect this at 417 00:25:14,359 --> 00:25:17,240 Speaker 2: a system level or did we just kind of execute 418 00:25:17,240 --> 00:25:19,520 Speaker 2: well on one goal, which is bring as much renewable 419 00:25:19,560 --> 00:25:22,640 Speaker 2: energy into the system, and now we're dealing with these 420 00:25:22,720 --> 00:25:27,120 Speaker 2: kind of downstream consequences, which are you know, the examples 421 00:25:27,160 --> 00:25:29,600 Speaker 2: you gave where now it doesn't always always operate as 422 00:25:29,600 --> 00:25:32,560 Speaker 2: we expected it to. It's creating other issues which is 423 00:25:32,640 --> 00:25:34,960 Speaker 2: driving us to rely back on fossil fuels again or 424 00:25:35,000 --> 00:25:38,280 Speaker 2: talk about LNG investments. And so what we look at 425 00:25:38,480 --> 00:25:43,359 Speaker 2: is optimizing the energy configuration as a whole, go through 426 00:25:43,359 --> 00:25:47,000 Speaker 2: the network lens, the generation lens, and you know, most 427 00:25:47,040 --> 00:25:50,879 Speaker 2: importantly the reliability and resiliency lens, and so what we 428 00:25:50,920 --> 00:25:55,000 Speaker 2: can do is simulate, you know, one ten thirty year 429 00:25:56,200 --> 00:25:59,360 Speaker 2: kind of outlooks, which is, all right, it sounds good 430 00:25:59,400 --> 00:26:01,720 Speaker 2: to put all this we here, but what is that 431 00:26:01,720 --> 00:26:03,639 Speaker 2: can look like five ten years from now? Is that 432 00:26:03,720 --> 00:26:06,400 Speaker 2: where the low growth is going to be other other 433 00:26:06,440 --> 00:26:08,880 Speaker 2: issues we're going to have around behind the meta assets, 434 00:26:08,920 --> 00:26:12,840 Speaker 2: whether it's more electric vehicle uptake. And we can start 435 00:26:12,880 --> 00:26:16,719 Speaker 2: to look at this on a twenty thirty year planning cycle, 436 00:26:17,400 --> 00:26:19,880 Speaker 2: where a lot of these decisions have been made on 437 00:26:19,920 --> 00:26:23,159 Speaker 2: a literally one week ahead basis, which is like, all right, 438 00:26:23,200 --> 00:26:25,440 Speaker 2: we're trying to get to this penetration rate of renewables. 439 00:26:25,840 --> 00:26:28,119 Speaker 2: Let's just build as much renewables as possible. Put the 440 00:26:28,160 --> 00:26:30,120 Speaker 2: next one here, the next one there, the next one there. 441 00:26:30,320 --> 00:26:32,959 Speaker 2: A look, we've got eighty percent renewables. That was a 442 00:26:33,000 --> 00:26:36,520 Speaker 2: noble achievement, but we weren't actually able to simulate what 443 00:26:36,560 --> 00:26:39,160 Speaker 2: that would mean for other consequences of how the energy 444 00:26:39,200 --> 00:26:42,760 Speaker 2: system works. And so that's what you know, again, the 445 00:26:42,840 --> 00:26:46,600 Speaker 2: risk of sounding to subjective. A behavior model enables you 446 00:26:46,640 --> 00:26:49,639 Speaker 2: to do is you can actually plan out millions of 447 00:26:49,680 --> 00:26:54,960 Speaker 2: scenarios and you can never obviously buy down every kind 448 00:26:55,000 --> 00:26:58,680 Speaker 2: of unintended consequence but get much closer to what does 449 00:26:59,520 --> 00:27:03,479 Speaker 2: a well oiled energy system look like across both network 450 00:27:03,480 --> 00:27:12,520 Speaker 2: and generation, you'll. 451 00:27:12,359 --> 00:27:14,480 Speaker 1: Probably think it's sort of quaint. But we just keep 452 00:27:14,560 --> 00:27:18,040 Speaker 1: arguing here about solar and whether we should encourage, you know, 453 00:27:18,119 --> 00:27:22,040 Speaker 1: through incentives like subsidies, more uptake of solar. You know, 454 00:27:22,040 --> 00:27:24,679 Speaker 1: America has been through this, Spain, Australia. You know, you 455 00:27:24,760 --> 00:27:28,000 Speaker 1: drive around the suburbs of Brisbane and places like that, 456 00:27:28,119 --> 00:27:31,720 Speaker 1: you see solar all over the place. Any any insights 457 00:27:31,720 --> 00:27:34,800 Speaker 1: from your time, particularly with fresh solar and in the 458 00:27:34,840 --> 00:27:37,639 Speaker 1: clean tech space about what you know, is this a 459 00:27:37,800 --> 00:27:39,960 Speaker 1: vible option for New Zealand what we need to do 460 00:27:40,080 --> 00:27:43,600 Speaker 1: to to get more capacity from solar online quickly? 461 00:27:44,200 --> 00:27:46,760 Speaker 2: Yeah. So it sounds like a bit of a contradictive 462 00:27:46,760 --> 00:27:49,240 Speaker 2: thing to say, just given how much time I spent 463 00:27:49,840 --> 00:27:54,080 Speaker 2: working in policy supported contexts. But my view has always been, 464 00:27:54,119 --> 00:27:57,359 Speaker 2: and this isn't just specific to sol or renewables, is 465 00:27:57,400 --> 00:28:02,320 Speaker 2: that you should never put in place a supporting policy 466 00:28:03,000 --> 00:28:05,680 Speaker 2: if there isn't a viable path to what you're supporting 467 00:28:05,720 --> 00:28:09,880 Speaker 2: getting to commercial viability without that. And so in markets 468 00:28:09,920 --> 00:28:16,400 Speaker 2: like Spain, the US, parts of Australia, some of those 469 00:28:16,440 --> 00:28:20,280 Speaker 2: programs that were you know, quite lucrative at first and 470 00:28:20,320 --> 00:28:23,760 Speaker 2: came under a bit of fire. Actually were the reasons 471 00:28:23,800 --> 00:28:28,240 Speaker 2: why entire industries invested in the soul of value chain 472 00:28:28,280 --> 00:28:31,040 Speaker 2: in those countries and got sold to a point where 473 00:28:31,200 --> 00:28:35,560 Speaker 2: it's now economically commercially viable without those subsidies. Now New 474 00:28:35,640 --> 00:28:39,520 Speaker 2: Zealand is not as much of a slam dunk argument 475 00:28:39,560 --> 00:28:41,480 Speaker 2: for all the obvious reasons. That doesn't have the same 476 00:28:41,560 --> 00:28:45,240 Speaker 2: level of sunlight, and so I think you'd have to 477 00:28:45,280 --> 00:28:48,600 Speaker 2: be very judicious around all right, what are we going 478 00:28:48,680 --> 00:28:51,520 Speaker 2: to do to support it, what in the economic value 479 00:28:51,600 --> 00:28:55,320 Speaker 2: chain actually can be moved forward. So one of the 480 00:28:55,360 --> 00:28:58,720 Speaker 2: big things that really unlocked the cost of solar globally 481 00:28:59,600 --> 00:29:01,520 Speaker 2: was the to financing because it was seen as a 482 00:29:01,600 --> 00:29:05,040 Speaker 2: risky asset. These policies and incentives, you know, gave people 483 00:29:05,040 --> 00:29:08,400 Speaker 2: the comfort to invest, and then everyone saw the operational 484 00:29:08,400 --> 00:29:11,120 Speaker 2: performance at all, right, these assets aren't so risky, So 485 00:29:11,160 --> 00:29:13,080 Speaker 2: are we going to finance it at a lower cost 486 00:29:13,080 --> 00:29:16,120 Speaker 2: of capital? So the cost of financing came down and 487 00:29:16,160 --> 00:29:19,120 Speaker 2: so that really achieved a goal. And so what you 488 00:29:19,200 --> 00:29:21,560 Speaker 2: know is always important is that you'd have to look 489 00:29:21,600 --> 00:29:23,960 Speaker 2: at what's the cost of solar in New Zealand right now? 490 00:29:24,680 --> 00:29:27,480 Speaker 2: How many of the inputs to that can actually change 491 00:29:27,800 --> 00:29:30,520 Speaker 2: because you can't actually impact sunlight, and sunlight is the 492 00:29:30,520 --> 00:29:34,320 Speaker 2: biggest input into the economic equation. Can the other inputs, 493 00:29:34,360 --> 00:29:37,280 Speaker 2: whether it's the cost to install a solar rooftop system 494 00:29:37,320 --> 00:29:40,160 Speaker 2: on a residential house because there's only one or two 495 00:29:40,240 --> 00:29:42,640 Speaker 2: people that do it right now, and scale drives cost 496 00:29:42,720 --> 00:29:46,280 Speaker 2: down the absolute cost of those systems. If there's enough scale, 497 00:29:46,360 --> 00:29:50,160 Speaker 2: can we procure you know, the system components at a 498 00:29:50,200 --> 00:29:54,120 Speaker 2: cheaper price. Is there a financing element to it? And 499 00:29:54,360 --> 00:29:57,480 Speaker 2: do we generally believe that if we kind of kickstart 500 00:29:57,480 --> 00:30:00,520 Speaker 2: this industry there is a path to become a bible 501 00:30:01,080 --> 00:30:03,640 Speaker 2: because there is a graveyard of industries. You know, the 502 00:30:03,680 --> 00:30:06,320 Speaker 2: auto industry in Australia is a good example, where for 503 00:30:06,440 --> 00:30:09,960 Speaker 2: decades it was subsidized. But the argument that Australia was 504 00:30:10,000 --> 00:30:11,760 Speaker 2: ever going to get to a labor point, labor cost 505 00:30:11,760 --> 00:30:14,160 Speaker 2: point where auto manufacturing was going to compete with China 506 00:30:14,680 --> 00:30:18,400 Speaker 2: or other markets was probably pretty aspirational. And so it's 507 00:30:18,400 --> 00:30:19,960 Speaker 2: a bit of a long winded answer, but just having 508 00:30:20,040 --> 00:30:23,000 Speaker 2: lived through, I kind of count them now. Dozens and 509 00:30:23,080 --> 00:30:27,400 Speaker 2: dozens and dozens of solar incentive policies. There is no 510 00:30:27,520 --> 00:30:30,360 Speaker 2: point going on the journey if you can't be confident 511 00:30:30,400 --> 00:30:34,080 Speaker 2: in the end of it. The industry and all the 512 00:30:34,120 --> 00:30:38,360 Speaker 2: ecosystem participants in it can play without the incentives because 513 00:30:38,400 --> 00:30:42,640 Speaker 2: otherwise you're just distracting from other things. They could have 514 00:30:42,720 --> 00:30:46,120 Speaker 2: a more commercially viable path to solving the same problem. Statement. 515 00:30:46,720 --> 00:30:50,400 Speaker 1: Yeah, and I think you know, the same equation needs 516 00:30:50,400 --> 00:30:53,880 Speaker 1: to be worked through for our heavy industries because we 517 00:30:53,960 --> 00:30:57,800 Speaker 1: may have eighty percent renewables, but overall, you know about 518 00:30:58,080 --> 00:31:01,840 Speaker 1: forty five percent of energy is not coming from the grid. 519 00:31:01,920 --> 00:31:04,720 Speaker 1: It's it's process heat, and you know, it's gas and 520 00:31:06,560 --> 00:31:10,760 Speaker 1: coal effectively, it's drying milk powder, it's it's log kilns, 521 00:31:10,840 --> 00:31:13,040 Speaker 1: all that sort of thing. So I guess you already 522 00:31:13,040 --> 00:31:15,280 Speaker 1: thinking about and have been with your customers about how 523 00:31:15,320 --> 00:31:19,920 Speaker 1: do we support the modeling of these physical assets that 524 00:31:19,920 --> 00:31:24,760 Speaker 1: aren't necessarily consistently drawing current from the grid. They're generating 525 00:31:24,880 --> 00:31:27,160 Speaker 1: energy in their own right, and how do we advise 526 00:31:27,200 --> 00:31:29,360 Speaker 1: them on the best way to do that sustainably. 527 00:31:29,760 --> 00:31:31,280 Speaker 2: Yeah, and I think that's part of it. So a 528 00:31:31,280 --> 00:31:33,360 Speaker 2: lot of what we look at now isn't do you 529 00:31:33,440 --> 00:31:36,000 Speaker 2: build more grid because to your point, a lot of 530 00:31:36,040 --> 00:31:39,400 Speaker 2: the assets that are approximate to the grid aren't necessarily 531 00:31:39,440 --> 00:31:42,600 Speaker 2: pulling down on current or electricity, and then it's like, 532 00:31:42,640 --> 00:31:45,480 Speaker 2: all right, how can we actually use assets like this 533 00:31:46,400 --> 00:31:49,760 Speaker 2: to complement the grid and relieve the grid? And so 534 00:31:50,080 --> 00:31:54,760 Speaker 2: a lot of the challenge around harnessing behind the meta 535 00:31:54,840 --> 00:32:01,040 Speaker 2: assets like batteries like solar rooftop PB is as a 536 00:32:01,160 --> 00:32:06,440 Speaker 2: massive I guess data opacity between what's happening behind the 537 00:32:06,480 --> 00:32:09,840 Speaker 2: meta and what the networks are seeing. And so what 538 00:32:09,840 --> 00:32:13,840 Speaker 2: that means that the networks place these export constraints because 539 00:32:13,840 --> 00:32:17,160 Speaker 2: they can't see what's coming in, and volatility is worse 540 00:32:17,200 --> 00:32:22,240 Speaker 2: than having energy pumped into the system. And there are 541 00:32:22,480 --> 00:32:26,080 Speaker 2: ways where once you start to connect the digital information 542 00:32:26,160 --> 00:32:29,720 Speaker 2: from what's having behind the meta to digital models of 543 00:32:29,800 --> 00:32:33,000 Speaker 2: the grid, you can start looking at these assets and 544 00:32:33,080 --> 00:32:37,200 Speaker 2: other generating assets as a creative as opposed to parasitic 545 00:32:37,960 --> 00:32:43,600 Speaker 2: or generating assets that you need to dramatically constrain because 546 00:32:43,680 --> 00:32:46,880 Speaker 2: you can't predict exactly what they're going to generate, and 547 00:32:46,920 --> 00:32:49,640 Speaker 2: having unpredictable energy pumped into a network is not a 548 00:32:49,640 --> 00:32:54,760 Speaker 2: good thing. And so I think this whole evolution around 549 00:32:54,880 --> 00:32:58,000 Speaker 2: the kind of pro humor so people who take the 550 00:32:58,040 --> 00:33:00,520 Speaker 2: generation and use of energy into their own hand hands. 551 00:33:01,080 --> 00:33:05,960 Speaker 2: But that kind of evolution is not being super well 552 00:33:05,960 --> 00:33:10,800 Speaker 2: connected to how the grid can treat those kinds of assets, 553 00:33:11,280 --> 00:33:14,920 Speaker 2: harness those assets, and then not constrain their ability to 554 00:33:14,920 --> 00:33:17,240 Speaker 2: contribute to the problem. Statement, Because the more that you 555 00:33:17,280 --> 00:33:22,000 Speaker 2: can put energy into people's hands beyond life centralized generating 556 00:33:22,000 --> 00:33:25,480 Speaker 2: assets in the network, it's actually a good thing. Provided 557 00:33:25,560 --> 00:33:28,240 Speaker 2: you back to what we're talking about earlier, the whole 558 00:33:28,280 --> 00:33:31,840 Speaker 2: thing is integrated and architect in a way that doesn't 559 00:33:32,160 --> 00:33:37,760 Speaker 2: disrupt reliability, doesn't disrupt grid performance, and doesn't create unintended 560 00:33:37,800 --> 00:33:40,360 Speaker 2: economic consequences, you know, somewhere else in the system. 561 00:33:41,360 --> 00:33:43,400 Speaker 1: So, Jack, you've raised I think you know, one hundred 562 00:33:43,440 --> 00:33:47,800 Speaker 1: and eighty million Australian over several funding rounds, now ninety 563 00:33:47,800 --> 00:33:51,240 Speaker 1: million in the latest one. What sort of the top 564 00:33:51,240 --> 00:33:53,480 Speaker 1: two or three things that you'll be able to now 565 00:33:53,600 --> 00:33:57,280 Speaker 1: do with this influx of capital from the likes of 566 00:33:57,320 --> 00:33:59,800 Speaker 1: TCV and the other equity partners that have put in 567 00:33:59,840 --> 00:34:00,360 Speaker 1: this round. 568 00:34:01,160 --> 00:34:03,600 Speaker 2: Yeah. Sure. So the biggest kind of focus is our 569 00:34:04,240 --> 00:34:08,560 Speaker 2: existing vertical energy So right now we're about eighty five 570 00:34:08,719 --> 00:34:12,319 Speaker 2: ninety percent market share in Australia but still very early 571 00:34:12,440 --> 00:34:17,799 Speaker 2: on our journey in the US, Europe, Asia Pacific. And 572 00:34:17,840 --> 00:34:20,040 Speaker 2: so really what this will enable us to do is 573 00:34:20,080 --> 00:34:24,759 Speaker 2: invest in more software engineers, more machine learning engineers to 574 00:34:24,840 --> 00:34:29,560 Speaker 2: really drive the platforms capabilities forward. There are problem statements 575 00:34:29,560 --> 00:34:32,480 Speaker 2: and energy that we didn't predict a year ago. So 576 00:34:32,560 --> 00:34:36,200 Speaker 2: the whole circular economy of data centers and AI compute 577 00:34:36,280 --> 00:34:38,839 Speaker 2: and the need for more energy and again, you know, 578 00:34:38,920 --> 00:34:41,759 Speaker 2: back to the example you gave just before, around you know, 579 00:34:41,800 --> 00:34:44,759 Speaker 2: these new assets that are both potentially dragging on the 580 00:34:44,800 --> 00:34:49,080 Speaker 2: grid but also generating, you know, to power data centers, 581 00:34:49,400 --> 00:34:52,000 Speaker 2: and so the whole energy space just keeps throwing up 582 00:34:52,040 --> 00:34:56,480 Speaker 2: really interesting challenges. That's the first focus. The second one 583 00:34:56,680 --> 00:35:00,520 Speaker 2: is starting to move into different infrastructure verticals, so starting 584 00:35:00,520 --> 00:35:04,520 Speaker 2: to look at areas like rail telecommunications. That's always been 585 00:35:04,560 --> 00:35:07,120 Speaker 2: an aspiration for the platform, and this will enable us 586 00:35:07,160 --> 00:35:11,280 Speaker 2: to really start taking that seriously. But really the primary 587 00:35:11,920 --> 00:35:16,319 Speaker 2: focus still remains our core vertical of energy and really 588 00:35:16,440 --> 00:35:19,240 Speaker 2: kind of going from what I would call the beginning 589 00:35:19,239 --> 00:35:21,960 Speaker 2: of a tipping point around understanding that if you're not 590 00:35:22,040 --> 00:35:26,239 Speaker 2: digitizing these workflows and solving these problems now, it really 591 00:35:26,280 --> 00:35:29,160 Speaker 2: is probably going to become industry unacceptable and making sure 592 00:35:29,280 --> 00:35:33,439 Speaker 2: that we're doing that in the most reliable, predictable and 593 00:35:33,680 --> 00:35:37,319 Speaker 2: you know, with everyone's adoption of AI, particularly for electricity 594 00:35:37,360 --> 00:35:40,480 Speaker 2: networks and critical infrastructure owners in a way where the 595 00:35:40,560 --> 00:35:44,040 Speaker 2: human engagement component is still very prevalent and they can 596 00:35:44,080 --> 00:35:46,919 Speaker 2: still see, all right, software is telling me to do this. 597 00:35:47,520 --> 00:35:49,600 Speaker 2: I can still take that back to the single line 598 00:35:49,640 --> 00:35:52,399 Speaker 2: diagram of the asset, what it's made from, and all 599 00:35:52,400 --> 00:35:55,359 Speaker 2: the engineering calculations that nailed me to decide that I'm 600 00:35:55,400 --> 00:35:58,920 Speaker 2: going to spend a dollar hardening Peter's pole instead of 601 00:35:59,200 --> 00:36:00,560 Speaker 2: spending ten dollars to replace it. 602 00:36:01,640 --> 00:36:04,239 Speaker 1: That's really intriguing what you're saying about that. You know, energy, 603 00:36:04,280 --> 00:36:06,520 Speaker 1: it's a sweet spot, but there are these other industry 604 00:36:06,600 --> 00:36:09,439 Speaker 1: verticals that you're looking at. We've heard and I've talked 605 00:36:09,480 --> 00:36:13,000 Speaker 1: to people, you know, digital twin evangelists for years saying 606 00:36:13,000 --> 00:36:16,520 Speaker 1: we basically need a digital twin for the country. And 607 00:36:16,719 --> 00:36:19,640 Speaker 1: they sort of did it after the christ Church earthquake 608 00:36:19,680 --> 00:36:22,279 Speaker 1: where they had to put together very quickly a lot 609 00:36:22,320 --> 00:36:26,640 Speaker 1: of underground infrastructure modeling and above ground stuff as well, 610 00:36:27,040 --> 00:36:29,680 Speaker 1: but then that was basically shut down after they'd rebuilt 611 00:36:29,760 --> 00:36:32,759 Speaker 1: the city. But the concept of having maybe an open 612 00:36:32,800 --> 00:36:35,520 Speaker 1: source model that everyone can access the electricity layer, the 613 00:36:35,560 --> 00:36:40,040 Speaker 1: transport layer, the utility layer. Is that are we any 614 00:36:40,200 --> 00:36:42,799 Speaker 1: closer to that becoming a reality? What are the sort 615 00:36:42,800 --> 00:36:45,640 Speaker 1: of the barriers still in front of us from achieving that? 616 00:36:46,200 --> 00:36:49,480 Speaker 2: Yeah, so that's the journey we're on now, which is 617 00:36:49,560 --> 00:36:54,640 Speaker 2: not just having one asset tied, but being able to 618 00:36:54,680 --> 00:36:58,080 Speaker 2: engage in the same kind of you know, analysis around 619 00:36:58,160 --> 00:37:02,239 Speaker 2: different assets, you know, to your point, show how they 620 00:37:02,320 --> 00:37:06,400 Speaker 2: all kind of integrate together. And you know, there's a 621 00:37:06,520 --> 00:37:09,600 Speaker 2: benefit in creating a digital twin to do something once, 622 00:37:10,280 --> 00:37:14,280 Speaker 2: so you know, to rebuild the infrastructure of a damaged city. 623 00:37:14,719 --> 00:37:18,920 Speaker 2: But the true goal of a digital twin, which is 624 00:37:20,400 --> 00:37:25,160 Speaker 2: what ours manifest, is it's an ongoing, dynamic, living, breathing model. 625 00:37:25,680 --> 00:37:27,959 Speaker 2: And so you know, we start with modeling ninety percent 626 00:37:28,000 --> 00:37:31,680 Speaker 2: of Australia's electricity network, but that's an ongoing breathing model. 627 00:37:31,680 --> 00:37:36,000 Speaker 2: And now we're looking to overlay rail and row and 628 00:37:36,200 --> 00:37:40,279 Speaker 2: other infrastructure because you know, it's very rare that you're 629 00:37:40,320 --> 00:37:43,240 Speaker 2: not going to see some crossover and how infrastructure behaves, 630 00:37:43,480 --> 00:37:46,359 Speaker 2: so particularly when infrastructure is underground. You've got water, you've 631 00:37:46,360 --> 00:37:49,600 Speaker 2: got gas, you've got electricity rail. A lot of the 632 00:37:49,640 --> 00:37:52,879 Speaker 2: issues in rail are very analogous to the issues in electricity. 633 00:37:53,440 --> 00:37:56,560 Speaker 2: Electricity is almost always playing a role in every other 634 00:37:56,600 --> 00:38:00,279 Speaker 2: critical infrastructure vertical. So yeah, that's the goal. The goal 635 00:38:00,400 --> 00:38:03,560 Speaker 2: is that you end up with a living, breathing version 636 00:38:04,200 --> 00:38:07,120 Speaker 2: of what's happening in the physical world. 637 00:38:08,280 --> 00:38:12,600 Speaker 1: Well, it's an incredible success with this company in a 638 00:38:12,640 --> 00:38:16,000 Speaker 1: relatively short period of time. Jack, congratulations on that, on 639 00:38:16,239 --> 00:38:18,279 Speaker 1: the recent capital rais, and thanks so much for coming 640 00:38:18,320 --> 00:38:19,200 Speaker 1: on the Business of Tech. 641 00:38:19,600 --> 00:38:21,640 Speaker 2: No, thanks so much, Peter, I really appreciate the opportunity. 642 00:38:26,560 --> 00:38:29,040 Speaker 1: That's it for this episode of the Business of Tech. 643 00:38:29,239 --> 00:38:31,680 Speaker 1: Thanks to Jack Curtis for coming on. We'll keep an 644 00:38:31,719 --> 00:38:34,879 Speaker 1: eye on nearer in the next few years. They've raised 645 00:38:34,920 --> 00:38:37,680 Speaker 1: one hundred and eighty million Australian in total so far, 646 00:38:37,840 --> 00:38:42,359 Speaker 1: so plenty of capital for the next chapter. And what 647 00:38:42,440 --> 00:38:45,839 Speaker 1: I'm really interested and intrigued to see is when they 648 00:38:45,880 --> 00:38:49,560 Speaker 1: start branching out into those other verticals like transport and 649 00:38:49,680 --> 00:38:53,640 Speaker 1: other types of infrastructure, and when we can, through open 650 00:38:53,640 --> 00:38:56,239 Speaker 1: standards and things like that, plug all of these into 651 00:38:56,320 --> 00:38:59,160 Speaker 1: each other so we get a real unified view off 652 00:38:59,200 --> 00:39:02,840 Speaker 1: the infrastructure that's so critical to the economy and society. 653 00:39:02,920 --> 00:39:05,759 Speaker 1: That's when it will get really interesting. If you enjoyed 654 00:39:05,800 --> 00:39:09,240 Speaker 1: this conversation, please follow the show, leave a rating or review, 655 00:39:09,360 --> 00:39:11,800 Speaker 1: and share it with someone who's thinking about the future 656 00:39:11,840 --> 00:39:14,840 Speaker 1: of New Zealand's infrastructure. You can find more of my 657 00:39:15,000 --> 00:39:18,400 Speaker 1: reporting on tech startups and innovation at Business Desk and 658 00:39:18,440 --> 00:39:20,799 Speaker 1: also in a New Zealand Listener, and as always, dropped 659 00:39:20,840 --> 00:39:24,279 Speaker 1: me a line with ideas for future episodes, things you'd 660 00:39:24,320 --> 00:39:27,040 Speaker 1: like to see discuss, people you'd like me to have on. 661 00:39:27,640 --> 00:39:29,919 Speaker 1: Thanks so much for listening and I'll catch you next week.