Machine learning and Artificial Intelligence in the Real Estate and Construction sector

Swapp, a British-Israeli ‘Contech’ start-up, raised 5.1 million GBP last week in venture capital. ‘Contech’ refers to the technology used in the planning, design and implementation process of large construction projects. Swapp’s offering is based around construction artificial intelligence (AI). Developers upload their plans, before receiving a selection of targeted planning options originating from an AI algorithm.


Public opinion surrounding AI has been divided since its inception. Traditional critics believe that whilst AI may facilitate greater data analysis and cognitive intuition, it lacks the human persona, which is integral to bespoke planning processes. However, enthusiasts maintain that AI represents a powerful instrument that complements an individual’s interactional skills, and that the harmonious adoption of these technologies will allow for greater internal and external operational efficiencies within the real estate and construction sector.


Both Point72, a hedge fund, and Entrée Capital, a venture capital firm, who provided the funding for Swapp, justified their investments by maintaining that Swapp were pioneering a new era of construction and real estate development. Swapp CEO, Eitan Tsarfati, believes that the company’s software will simplify the current planning process. Upon uploading interior and exterior preferences for the structure and design, customers are set to receive an assortment of algorithmically induced planning proposals that aim to not only minimise cost inefficiencies, but also offer a more holistic package to the developer.


Incumbent property development proposals require the profusion of calculated design, planning, and financing considerations. This typically takes time and a phalanx of professionals. AI technology offers the chance to streamline the multi-dimensional nature of development conception, allowing developers to meticulously plan, control and visualise each phase, readily adjusting the budget in an ad-hoc style to optimise schedules. Visual AI is crucial for monitoring any deviations from the construction blueprint and is proficient in illustrating multiple outcomes with high precision and accuracy. Machine Learning (ML) is particularly useful when forecasting variable construction issues as it uses advanced pattern recognition and can predict external factor implications caused by bad weather or supply chain malfunctions, for example. This is particularly vital for corporate contingency planning. The technology is capable of making autonomous and holistic decisions, largely free from human myopia or mistakes. If certain properties are selling faster, the platform leverages this information and recommends appropriate design strategies that align with market demand to maximise sales revenue upon completion.


Another interesting area as far as investors and development are concerned is ‘property intelligence’. Through autonomous robotics, drones are already capable of surveying and completing spatial land valuation. In conjunction with predictive and sensitivity analytics that draws from comparative data pools, the lengthy process of procuring planning permission could be shortened to a matter of days. This will allow buyers and investors to gauge whether an investment is optimal and help to reduce tenant vacancy within both the sales and leasing markets. In a study carried out by Professor Thies Lindenthal from Cambridge University on sales price premia over architectural styles, it was deemed that ML classifiers were at least as reliable, if not more so, than human counterparts for mass appraisals on both a building and neighbourhood level. This is indicative of the scale on which artificial intelligence can aid incumbent professionals, enabling them to further their understanding of architectural styles and the impacts this has upon valuation.


Moreover,  AI offers a more creative and authentic outlook towards property design and facilitates clear valuation outlines, while also offering a replacement for time-consuming underwriting and research roles. For example, the American technology company, Skyline AI, uses predictive analytics to assess property value by accessing pools of data from 130 different environments.  Even basic AI is capable of factoring in intricate details such as broadband signal capacity or YELP reviews for local restaurants. The instantaneity and thorough due diligence that AI is capable of saves the realtor research time, freeing time for these individuals to focus more upon the interpersonal aspect of selling a property.


Overall, COVID-19 has played a key role in expediting this transition towards AI, given that many companies and industries must adapt to survive. It would appear clear that the use of AI will continue to proliferate in the coming years once corporations are familiar with the complementarities that AI has to offer, from assistance with the initial design conception, ultimately through to the final property valuation.


By Tristan May

Sector Head: Theo Thomas