Digital Quality Control of Real Estate
The ETH-Spin-off Archilyse has developed a digital solution that captures the characteristics of buildings and architecture and processes them in interpretable data and simulations. The aim is to assess the quality of floor plans in specific locations to quantify it and to make it comparable. The idea has the potential to set new standards in the whole real estate industry.
Nowadays digital tools measure everything. They track the state of health or fitness and identify any need for action. Why should not this also be possible for existing or planned architecture? This must have been the question that Matthias Standfest started his research with and which, a few years later, led to the founding of Archilyse. After his studies of philosophy and architecture, the native Austrian, wrote his doctoral thesis at the ETH Zurich. His focus was the assessment of architecture with machine learning. In his dissertation he developed algorithms which were further evolved by Archilyse and are now available for practical applications. Apparently the offer responds to recent trends. Major Swiss investors are on board and Archilyse already has big players of the local real estate market as customers.
Input: Floor Plan & Address > API Output: Data & Simulations
“Our core competence is the API,” explains Margarete Sotier. The native German is member of the Archilyse Founding Team and is responsible for public relations at the company. She is the interface who bridges the gap between abstract technology and interested potential users. The I in the word API also stands for the interface (Wikipedia translates API with “interface for application programming”). With its API, Archilyse is able to process a collection of information into a coherent real estate assessment. The company calls this achievement a “holistic analysis and evaluation of architectural quality”.
Input: The data and information exchange is made online. Interested companies deliver floor plan data (as a 2D-PDF, image file or an ifc BIM model) together with the address of the object. In addition to this basic information, additional information can be provided: room height, rental price, floor space, year of construction, etc. Archilyse then enriches this with internal input like information on the location and preferences of potential users.
Processing: Based on this information the Archilyse tools creates a basic 3D-BIM model which is then placed in a 3D of the surrounding environment. Thanks to the artificial intelligence of the tools, this “build-up process” is accelerating with increasing order numbers. “Initially, this process took about half a day,” says Margarete Sotier. “We’ve already reduced that time to a few seconds.” Based on this digital simulation setup, the analysis process starts. Numerous properties are determined: view, daylight conditions, noise pollution etc.
What makes Archilyse unique is that the analysis is enriched by research results that are included in the algorithms that analyse the floor plans. Thus, Archilyse collaborates with well-known research institutions such as ETH Zurich, University College London, the Royal Academy of Art and the Bauhaus University Weimar. This allows the API to identify how attractive or unattractive the floor plans are for different “types” of users. “Types” in this case mean people of different ages in different household configurations: families with children, elderly people, shared flats etc.
Output: The analytical results that Archilyse delivers are reports enriched with various simulations. Archilyse’s free open source front end includes a number of exportable data visualisations. Those help to identify and discuss complex relationships intuitively and to take decisions based on facts.
The report includes location, view and architectural simulations. The location simulation, for example, shows the distances to the nearest hospital or the nearest electric filling stations. The visual simulation determines if there is a view on greenery, waters, etc. from a specific point in an apartment. The architecture simulations can check functions, geometries or compliance with norms and standards.
In fact, the tools make real estate quality quantifiable, measurable and comparable on the basis of numerous precise parameters. The gut feeling is now reassured by the competence of the API solution.
But where does the data that Archilyse uses come from? Archilyse uses open source data such as Open Street Map and buys topographical data. The analyses are currently temporarily limited to the D-A-CH (Germany, Austria, Switzerland) but is globally applicable and initial inquiries from Asia, Scandinavia and the US accelerate the globalization of Archilyse.
spaces (2), visible volume from standing height (3); (bottom left): total visible volume from seat height
(1), acoustics simulation(2), visible volume from seat height (3). Volume large (red) and low (blue). (Foto: Archilyse)
Different Applications For Different Use Cases
The offer of Archilyse creates quite an upheaval in the market. The start-up has already realised a pilot project with the Swiss Life and PriceHubble. There it could be proved that the architectural quality is directly related to the rental income. «For the apartments that Archilyse certified with high quality, the revenue was high», states the founder and CEO, Matthias Steadfast in an interview with Swiss Architects.
The determination of quantifiable quality criteria was also the focus of a Credit Suisse publication on the Swiss real estate market in 2019 with the title: «Location, location, floor plan». It relates to the often heard answer to the question of the three most important factors for evaluating a property. This was so far: location, location, location. In the article titled “It’s the floor plan, stupid!” (Based on Bill Clinton’s instructive campaign slogan “It’s the economy, stupid!”) it is explained that intelligent analytical tools and algorithms can now objectively assess various aspects of floor plan quality. The assessment criteria used include the flexibility of use and the connectivity between the different rooms of a residential unit. Archilyse also systematically investigates the room’s ability to be furnished — with its algorithms, a rectangle of any size (e.g. a double bed of 200 x 200 cm) can be tested for all possible interior variations in the room. Conclusion of the furnishing analysis: Square-cut rooms can be furnished more diversely. Ergo, they deliver quality.
Quality: Quo Vadis?
“We measure real estate quality” is one important slogan of Archilyse. “But the interpretation of our data is outsourced,” says Margarete Sotier in a conversation about the company’s contribution to the evaluation of the delivered data. Nevertheless, one must take the claim seriously and deal with it — especially since the possibilities of the API are strongly demanded.
The team of Archilyse could already prove with their evaluations there is more to get out of certain properties. Some architects may freak out with the idea that quality can be assessed by the digital analysis of individual floor plans. Aren’t there haptic elements, smell, the general urban planning and many more factors that should be considered beyond the housing or office floor plans?
With Digitisation and new technological tools a battle for the architectural sovereignty of interpretation has started. This challenges previously respected competencies of entire professional groups. This must be seen in the realm of the global nature of digitisation.
