At Planalogic, the designer doesn’t have to keep changing the parameters each time; within defined limits, the computer does this using algorithms and artificial intelligence, generating millions of design solutions for each project. In generative design, as it is called, the client can then select the best or most suitable options. ‘In the projects we have delivered so far, an efficiency gain of 10% to 20% has been achieved,’ says Hugo Jager, a partner at Planalogic. This might be anything from projects for dozens of homes, to urban locations with high-rise, mixed-use buildings and a thousand to 1,500 apartments. For the digital platform, it makes no difference.
According to Van Wylick, the challenges facing the construction and real estate sector are big and the assistance of data and artificial intelligence is indispensable ‘There’s a huge housing shortage and the sector has a major climate impact. 36% of CO2 emissions come from the built environment. Eight of the United Nations’ Sustainable Development Goals apply directly or indirectly to real estate and they are all interlinked. If you build more energy-efficiently, something else has to be sacrificed. Or it becomes less affordable.’ A computer can find the optimum balance, but a human can’t, is what he is saying.
‘The way we learned to work is perfect for a world that is disappearing,’ says Van Wylick. He is talking about the typical development process. An architect is hired, a municipality has to give its blessing, the design is worked out, the construction costs turn out to be too high, the design has to be modified again, the final go-ahead is given and only then can the builder start work. ‘It’s labour intensive and, because of this, takes a really long time. Those two things make it expensive and you only have one, or a handful of alternatives. We take a smarter approach. With generative design, we generate millions of results, with not much work and in a short amount of time, on the basis of which the client can make an assessment.’
According to Jager and Van Wylick, a process which normally takes four months can be shortened to four or five weeks. And there are literally more options. 'In one housing construction project, it was found that, with one design option, for just €16,000 more in construction costs, the number of square metres increased from 186 m2 to 222 m2. This isn't a commonly used surface area but, it was possible to offer a unique and attractive product at such relatively low construction costs.’
The programme calculates other factors too. How many homes fit on a development location, how enough solar panels can be fitted on the roofs, whether the residents can see into their neighbours’ back gardens and how the windows can be positioned so that hardly any energy is wasted. Or, as was recently calculated for a project, whether the neighbourhood is so energy-efficient and generates and stores so much energy that it becomes entirely self-sufficient. Jager explains, ‘The neighbourhood can be built off-grid.’
Joosen expects that efficiency can be improved in other areas along the chain. ‘Right now, a design is converted by the builder to BIM (the 3D model which Joosen mentioned earlier, ed.) and buyers and estimators search for the right material for the right price. But with parametric design, you can directly link to the steel manufacturer who custom-cuts the steel pipes for the stadium, for instance. The only thing the manufacturer needs is an excel sheet with numbers. You can use those numbers like a construction kit during assembly on the construction site. Much of the drawing work can be automated. In nine out of ten cases, you can simply send through a data package.’
Read the original Dutch story on the website of Het Financieele Dagblad.
Copyright reserved Het Financieele Dagblad.