Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
AI’s fields of application are multiplying in the real estate value chain. Identification of potential renovation or construction sites, creation of thousands of possible building configurations, predicting accurately market trends. Beams and doors recognition from the cloud to create a BIM model, hazard identification on a site examined using videos etc.
Major property owners, REITs and property management companies sens the ongoing evolution in the industry. As real estate is the largest asset class in the world, stakeholders from adjacent sectors are getting involved. Insurance companies, energy companies, banks try to interface with this new PropTech startup wave.
Energy efficiency management is nowadays an area that particularly mobilises this technology. An AI application, developed by Ubiant and installed by Engie in 140 schools in the City of Paris since 2016, optimises consumption according to class size, number of students, temperature, etc. Completed by renovation work on buildings and boiler rooms; it will save 30% energy and reduce greenhouse gas emissions accordingly.
For the residential real estate, part of the marketing and sales effort is being enhanced through data-driven approaches. Robots already prewrite many real estate ads, whiles customers may use chatbots during their journey. Data Science techniques help to better compute the probability of a potential customer buying an apartment; according to their info, their interactions on the site, the apartment’s data, cookie scoring from third party supplier, etc.
As for the mobile phone in its early days, likely, we do not suspect 80% of the potential of smart buildings. The automation of building functions emerged in the late 1980s: motorised shutters, gates and garage doors, for example. But the recent increase in the number of connected objects and the integration of AI into the building marked a turning point. From now on, the idea is to transform buildings into service platforms through AI.
The first level of AI is an automated system that defines possible cases and associates them with precise action. Such as turning the lights off when the room is empty.
The second level is machine learning. Here, part of the algorithm is programmed, but another part improves by itself. It gradually integrates data and proposes adjustments on a case-by-case basis. Legrand thus imagined that an AI would control the green signal indicating the exit in the event of a fire. Adapted to the information on the smoking areas, the building plan, to indicate the right direction.
Finally, the third level is the usage of reinforcement and deep learning; possible through the integration of a large amount of curated data. The models generate outputs automatically, i.e. without prior programming. For example, for the recognition of a human face, there is no need to program its description. The AI will itself draw a diagram, after “digesting” a large number of images of this face, to define and being able to recognise the patterns, as intercom systems equipped with cameras already do.
In a few years’ time, the connected and adaptive building will be a necessity, an obvious one. In real estate, there are yet to be any revolutionary discoveries, as it is the case with mobility through autonomous cars. During this period of expansion, most of the Real Estate players slowly experiment the relevance of new approaches to their core business.
To integrate AI, construction players rely both on the plethora of start-ups and the giants of this technology such as Google, Amazon or IBM. All switches, sockets, shutters, connected door openers can be interoperated – i.e. linked to another company’s services – such as Alexa or Google Home. This interoperability of equipment, which is not yet guaranteed, is necessary for a building to be able to adapt to the rapid progress of technology and enable the vision of a Smart City.
It is already foreseeable today that AI applications will be interlinked with almost all applications. It will become an indispensable part of everyday life. Are we going to accept to live in an apartment full of sensors, fed by our data? Significant obstacles remain from the end-user adoption. People rightly fear data leakage in the comfort of their home.
From a technical perspective data availability, standardisation and usage due to inconsistent data models and legal requirements is a massive challenge. As a foundation for efficient AI use, improved data transparency and standardisation is critical to the real estate industry.
arrows & pagination