Hello, I'm here to present the autonomous engineer, she's trained to reduce risky cost and carbon in real estate development projects. That's what we need today, isn't it? So, to give you an example of an industry that has moved from having humans doing the engineering to machines, this is a component from the automotive industry in the real estate industry; we're all doing on what we see on the left-hand side.
So, everything that is engineered in a building like this is done by humans, or used to be, and we're moving that industry towards that left-and-side. For this component, there is a reduction of 70% in material usage. The real estate industry uses 40% of all the material in the world.
So, I think we need to question ourselves; can we really allow humans to do the engineering going forward? We also have a situation where the high complexity of doing the engineering of a building is not something we easily solve with human brains. So, we do stuff like this for no apparent reason: we shift the AA 180° in these ducts, causing a pressure loss that is six times as much as if you have taken a 45° angle out to that diffuser.
Instead, and then saving six times as much energy. And you will also notice that that fire detector is probably the last one to detect the fire in that building. There are certain rules on the distance between fresh air and water and a fire detector, but the high complexity in getting everything right is too complex for humans.
Everything is known, so it's a sweet spot for AI, and that's what we're utilizing in conely. So, we have called her the autonomous engineer, because we're real estate first, we're Engineers first, and then we use AI to solve explicit problems that we have when we do the engineering in buildings and infrastructure. So, going to give you some concrete examples to see how the machine can outperform human teams.
This first one might look like it's very close to architecture; we're Engineers, and architecture is a different education, different trade, different industry. But shuffling around standardized rooms, standardized apartments, is pure mathematics, and I believe Engineers are better at mathematics than Architects. And this is a way to get more value on your buildings.
So, in this case, for instance, it's a commercial building in Oslo, it's going to be turned down, and there's a hotel coming up. Hotels normally consist of standardized rooms, right? So, we asked the machine to find a configuration that gives most value, most tiny Norwegian kroners per night for the owner, and in this case, the machine found 24 more rooms than a manual team was able to do.
24 more rooms in this hotel means 7 million US dollars extra worth of value for that developer. It's a major increase of value within hours. Another place we see her performs really well is in all the technical systems.
So, if you have a geometry from an architect of a building, placing all the technical components in the ceiling grids, connecting it with ducts and pipes, and cable trays, is actually a huge puzzle of known physics; we know the requirements; we know the regulations; but it's really really complex. So, again, perfect for AI, not so perfect for human Grooms like we do today. And you will see here - this is a real project in Norway - where you normally use 6 to 12, sometimes 15 months with a large team to get the whole engineering in place.
In this project, we spent two two days. It was a startup meeting with a customer; we put in all the algorithms, let them go, and we did a quality assessment. Then we delivered to the customer, so it took it in this case from 12 months to 2 days to deliver the full ceiling grid.
Ready for tender for this developer, there's a massive amount of savings for man-hours; it's a massive amount of saving for money in that phase where you cannot have tenants in your building and get income. Then we have another part of the engineering; we have these plant rooms in today's says nobody is actually responsible for the size of this plant room. And it takes a lot of space that you could either sell, you could have tenants renting it, or you could avoid building it.
So, just taking mathematics from the shipping industry, we solve these plant rooms, so everything fits there; it's situated so it's easy to operate; it's a good place for the operation-maintenance staff to work, and then we can have the right size of these plant rooms. We normally reduce these rooms with about 50 to 20% in space, and that is a lot of extra worth for the developer and also a lot of cost savings in operation and maintenance because these rooms cost to operate and maintain as well. So, what I've shown now is pure engineering tools, but of course, an engineer would normally also hand-handle a lot of documentation in this industry; it's very documented focus.
So, we made some tools that can take out risk in huge packages of documentation as well. This was a very hard tool to explain to your customers before; it was a lot easier to explain this natural language processing technology after speech came around. So, but we'll basically take the whole tender documentation package, add it in, make it into data, train the data models, and we'll look for risk.
If we find risk, we tell the project manager where you have risk, so you can fix it before sending it out for tendering. Normally, a package like that is hundreds of documents; in this case, it was 650, 53 documents. Such fast for the machine to go in and look for risk, discrepancies, references to outdated standards, everything that should be sorted out before sending it for tender to take down the risk, and then the cost of course, and less problem in the development projects.
We also do this with a handover documentation; so when you get a building, if you um, developer and you get a building, you also get tons of documentation for that building; everything is described and that is in PDFs. So, in the handover, we simply take all that documentation, make it into data, and we look. Is everything there?
Is something missing? Is it the right quality? Have they tried to written something so the warranty doesn't count?
Everything and we sort the documentation, prepare it for the OMM system; retrieve all the tasks that you need to set up for OMM system so we make it ready for operation to take over that building, and then the funny thing, again, with the speech, we built on Microsoft Asher, so we had access to take the speech technology into our platform as well because we sat there when it came around. We already trained data models, so we just applied the technology, and then allowing all our customers to talk to their buildings, so instead of having a graphical digital twin, we sometimes say, we kill the digital twin in real estate; you can now talk to your documentation; instead, and that's what actually the operation and maintenance need to operate that building. So, we were lucky; they came along, and a lot of AI tool that we could just utilize, although we were building a lot beforehand, and I think that's a great value of having a tech company today, there's so much technology being developed, so it's getting easier and easier to take these really valuable tools out to uh, customers.
A little bit about our business, um, I founded the company in August 2020, in Oslo, in the middle of the pandemic, um, we have now scaled so we have more or less the whole private Market in Norway. The only problem is that they don't develop that much, but luckily we also have the public market H and they do develop. They build schools, hospitals, kindergarten, social housing, etc.
So, it's a good market for us. But as we saw, the private Market was going down; we hurried to UK, so we were able this year to land in the UK, and that's also where we have the first office outside of the Nordics, and then engineering of buildings is actually something you can scale globally because you have national standards, the vendors are Global, so it's a good opportunity to build a tool that also scales globally, so that's what we're working at at the moment. In total, we have raised about 4 million US dollars we have uh no VCS at moment; we just have industry investors and Inberg and I still have control.
Oh, we expect Revenue this year to land on about uh 2 million US dollars; still quite small numbers but last year, we only have a little bit more than a half, so it's a good growth, um, and uh, we will probably wait until Q1 2025 to to raise the next round. Because we want the numbers to get higher, and we see that we're close to getting bigger numbers, and that is um because we are also going into uh, US uh, we have landed in Japan, since we made a slide de, we landed in France, uh, we're going into uh, the Middle East, and um, we plan to scale to the global market as quickly as we can. We have a highly diverse uh Team, we have 16 different nationalities sitting in Oslo.
A large team of people with PhDs in various um, fields of AI and Mathematics and physics, and uh, we are female-founded and led, so if you're interested in uh, talking to us any further, scan this QR code and reach out on LinkedIn, and uh, we very happy to follow up. Thank you.
