Digitalization here, artificial intelligence there – neither means anything without sufficient data. And what can data actually be used for in practice? At the PropTech Innovation Now event on May 16, 2024, real use cases provided the answers.
The late-afternoon, final concert hall section of PropTech Innovation Now focused on conscious data management and the practical application of the many advanced technological solutions presented throughout the day. More precisely, it was about what we can already deploy right now, even without significant investment, because services, companies, and datasets are available to support more effective decision-making.
“I have a manual or description for an air-handling unit, but I only need a certain part of that information, yet I am facing a 200-page document,” said Marcell Laczkó, AI Consultant at AI Partners, outlining a hypothetical but very realistic situation in his presentation. The solution? Naturally, generative AI. “I tell it what I need from this – in this case, a simplified extract from the user manual – and I get it generated.”
The same process applies when we receive a long, hard-to-interpret ESG report. “I am a property manager, a sales director, a CEO, or I am the designer of this building. What in this ESG report is relevant to me?” AI takes on the text and, taking into account the questioner’s perspective, summarizes the relevant information. According to Laczkó, this is no longer a dream but reality. Freely available large language models and generative procedures can even help those companies and real estate players who otherwise ignore technological development.
In Laczkó’s fourth example, he presented a custom AI procedure that helps summarize often very complex and lengthy lease agreements. This not only saves time but also gray hairs, as sifting through thousands of pages of legal text is hardly relaxing. Yet the data is in the text – it just has to be found.
Marcell Laczkó, AI Partners (Photo: Glódi Balázs)
Just like the areas where we want to build. In the past, on-site inspections were needed to understand a location. Nowadays, we have tools such as satellites or LiDARs, which help map, model, and analyze our environment down to the last square centimeter. Tamás Tomor, co-owner and business development director of Envirosense, a geospatial solutions provider, presented practical procedures based on this.
“In 2021, we started a nationwide aerial LiDAR survey in Hungary. This survey covers the entire country,” he said, then presented the resulting geometric dataset. Imagine a gigantic digital map built from point clouds, covering everything from the first wooden shed to the last bush. Because this is an enormous amount of data, artificial intelligence helps classify the mapped elements, identifying exactly what objects each cluster represents. The project is currently at 60 percent, and Tomor’s team estimates another two years of work will be needed to map the whole country. Building modeling, urban planning, ESG analysis, investments… such a database could be useful in countless areas, even for enhancing livability.
“Green spaces within municipalities are becoming increasingly important, especially trees. Livable environments are becoming more valued; we feel better in green surroundings than among barren stones,” said Tomor. Yes, but maintaining green areas is costly for municipalities. To prepare properly, they first need to know exactly where green spaces are, what trees grow there, and how large they are. The Envirosense project provides all this – and that’s just one use case. With the right data, environments can be tailored to individual needs.
Of course, this works the other way around too: geospatial data collection can reveal how existing buildings affect their surroundings, each other, or even residents. Béla Szivák, managing director of Paulinyi & Partners, presented a project conducted in cooperation with the European Space Agency to study predictive urban insulation using satellite data.
“This is the map of Milan,” Szivák said, pointing to a thermal aerial photo glowing in red and gold. “On the right, in the corner, you can see the impact of a massively built-up shopping center on its environment” – he pointed to a deep red, dark spot clearly radiating heat across the entire district. Being nearby is far less energy-efficient than, say, living next to a park. Such data helps determine the insulation and cooling solutions appropriate for given sites and sets an example for newly built projects.
Béla Szivák, Paulinyi & Partners, during the informal discussion at the end of the event (Photo: Balázs Glódi)
Of course, not all data is the same. “We heard a lot about how to collect scientific data and optimize it, but not a word about collecting social science data. For me, this is a huge gap, because a building is not built just to exist, but to be used,” said Viktor Hegedüs, CEO of APFM-Systems, a company specializing in proptech solutions, at the closing roundtable on data science moderated by Zoltán Kalmár. “If we completely leave out the constantly changing motivations and behaviors of the people inside, then we may know how to operate a building – but one with no one in it.”
According to Gyula Győri, Facility Management Director at CPI, it is already a problem that company leaders consider innovation only as a tool for increasing cost efficiency, while overall quality improvement is lost in the pursuit of revenue. “I believe that the search for solutions will bring digitalization closer,” he explained. “Even if I look only at our own operations, we need to provide so much data to different clients for daily operations that it is no longer achievable with traditional methods.” Overall, it is the limitations of established practices that push companies toward smart data use – but much progress is still needed to ensure that a purely technocratic approach does not ignore individual needs.
Roundtable on conscious data use with Zoltán Kalmár (PropTech Hungary), Gyula Győri (CPI), Dr. Viktor Hegedüs (APFM-Systems), and Tamás Giller (Neops) (Photo: Balázs Glódi)