From Bedrooms to Biometrics: How AI is Changing the Real Estate Sector
A dive into the new methods of real estate valuations using AI
Use of AI in Real Estate
In the world of real estate, there have been a significant increase in the usage of AI. Based on work by the McKinsey Global Institute (MGI), it is believed that generative AI could generate $110 billion to $180 billion or more in value for the real estate industry.
Real estate has also been more conversative historically on the adoption of new technology. As such, there are still a lot of opportunities waiting to be uncovered and more importantly — implemented.
While usage can range from reviewing and expediting lease documents to introducing AI visualization for home tours, in this article, we are going to look into potential changes in how a property is valued.
The Innovative Approach in Real Estate Valuation
It is commonly agreed that allocating between 25 and 40 percent of your net worth to real estate (including your personal home) is a fine balance to your total portfolio. This also means:
The value of the real estate that you own is a significant portion of your total net worth.
In this sector, where the value of a property often hinges on how many sinks there are in the kitchen, researchers at University of Colorado Boulder have been looking at how AI can be used and bring a fresh light on how properties are appraised and valued.
This new method combines the ‘hard facts’ — such as the number of bedroom, bathroom, and size of the background of traditional valuation with the ‘soft touches’ of modern computer vision.
Valuation of real estate often combines many various factors, such as neighborhood, market conditions, home features, upgrades, etc. The new method taps into the subtler aspects of a property — the kind that makes you say, “Wow, this is what I call a “home!”
Using self-supervised vision transformers, this method interprets images to extract the unquantifiable: the charm of a living room, the allure of a well-kept garden, and even marble on the kitchen island.
This new method a breakthrough in computer vision and deep learning in the past year or so. The proposed algorithm combines machine learning and computer vision and also pricing models trained on real estate data to estimate the value of a given property.
The results show that the proposed algorithm outperforms traditional appraisal methods that do not leverage property imaging. With improvements of training data and algorithm, the methods will become more powerful in its efficiency and accuracy of property valuation.
What It Means for the Real Estate Sector
To recap, with the use of AI, there can be significant improvements for appraisals and also other day-to-day tasks in the real estate industry.
Revolutionizing Property Valuation
- Holistic Property Assessment: Traditionally, property valuation has leaned heavily on quantitative data like size, location, and age. Even though the industry have used these “hard-and-fast” reals to determine the value of a property. Technology brings a paradigm shift by incorporating qualitative aspects. With It can analyze visual cues to assess aesthetic appeal, upkeep, and even neighborhood ambiance, giving a more rounded view of a property’s worth.
- Accuracy and Objectivity: Humans are naturally subjective, but AI does not take into consideration whether the wallpaper doesn’t match a specific individual’s taste — it’s all about the data. This technology reduces human bias in valuations, leading to a more objective and consistent appraisal process.
- Efficiency in Appraisals: Manual appraisals are time-consuming. This AI-driven approach can process visual data at lightning speeds, potentially reducing appraisal times from days to hours or even minutes. This efficiency could revolutionize turnaround times in sales and lending processes.
Beyond Real Estate: Diverse Industrial Applications
There are a wide range of use cases for this type of technology:
- Retail and E-commerce: This technology could potentially assess the popularity of a product sold in retail — like a “vintage jacket”. Or in E-commerce, whether a certain mock up picture is more attractive than the other when displaying it on your Shopify store.
- Automotive Industry: Vehicle valuations can go beyond evaluation based on the exterior, driven mileage or mechanics. More factors like interior designs or customizations can come into play.
- Insurance: In insurance, assessing claims could even be more efficient and accurate with just images taken at the scene.
- Hospitality and Tourism: Hotels and resorts can now potentially adjust their room price based on the “view” of the beach.
- Urban Planning: For city planners, machine learning, computer vision could be incorporated with municipal planning to assess and come up with the best locations for schools or hospitals.
Future Possibilities and Challenges
As amazing as this technology sounds, it’s not without its weaknesses. The challenges of ensuring ethical data use and privacy remain paramount and critical as we move forward with newer (and better) technology every single day.
Let’s also not forget, while AI is intelligent, human experience is still critical in any tasks we do. AI is not meant to be a replacement, but industry professionals should learn how to best use this new technology to help them deliver better, and more accurate work.
Conclusion
With the groundbreaking work from the University of Colorado Boulder, we’re stepping into a future where property valuation isn’t just about numbers and square footage.
This novel approach is set to redefine how we look at real estate valuation and beyond, proving that sometimes, the best way to see the value is to look at the big picture, literally.
References:
- “Real Estate Property Valuation using Self-Supervised Vision Transformers” by Mahdieh Yazdani, Maziar Raissi, University of Colorado Boulder, 2023.
