Pi Labs research reveals that AI could reduce the annual carbon footprint of the built world by 50% by 2030

Global proptech venture capital firm Pi Labs has today launched ‘Sustainably Intelligent: AI for a Greener Built World’, a research report that uncovers new details about the transformative impact that artificial intelligence will have on reducing carbon impact in the real estate industry and built environment.

The research was conducted by Luke Graham, head of research at Pi Labs. He joined Pi Labs from the University of Oxford’s Future of Real Estate Initiative where he continues as a lecturer. In 2023 Luke was selected to chair the UK Housing Minister’s PropTech Roundtable.

An industry-wide green transition is currently underway to halve emissions by 2030, and achieve net-zero before 2050, the goals necessary to secure the future health of the planet, according to climate scientists. With the built world responsible for 40% of overall emissions, it is frequently identified as critical to the planet’s decarbonisation drive. The new report from Pi Labs explores the role that AI could have in achieving this.

For the first time, the research team at Pi Labs has calculated the carbon reduction potential of AI to reduce the built world’s environmental footprint. The team found that 5.81 to 6.46 gigatonnes CO2-e of greenhouse gas (GHG) emissions can be avoided annually by 2030 with industry-wide adoption of just four AI use cases. This would offset the entire annual carbon footprint of the United States—the world’s second-largest GHG emitter – according to 2022 figures.

After examining 68 sustainability-oriented use cases for AI across the real estate value chain, the following were selected to conduct the analysis with the use of proprietary and public data; reducing raw material inputs through generative design, preventing construction rework with 3D AI analytics, reducing building energy intensity with AI-enabled smart building technology, and redirecting demolition waste with AI-enabled waste material analysis.

Generative design: researchers at ETH Zurich have deployed 3D printing to reduce construction material inputs by 70 percent, as previously reported by WEF. During the research for Sustainably intelligent, Pi Labs identified early-stage startups deploying AI to design such structures at scale.

Construction rework: In 2030, Oxford Economics projects global construction output to be $15.2 trillion in 2030. Academic research indicates 2-30 percent of construction expenditure is wasted through “rework”. Companies such as Contilio deploy LiDAR and 3D AI to identify defective work early and prevent rework on construction sites, which means preventing the use of excess construction materials.

Building energy waste: energy waste in buildings is a well-known global phenomenon. Heating and cooling systems running concurrently occurs regularly in commercial buildings. AI-enabled tech start-ups such as Demand Logic have successfully deployed data analytics to reduce energy waste in buildings. 

Demolition waste: waste sorting remains a manual and inefficient process in many markets. Computer vision technologies such as Sorted.io are being deployed to improve the economics of waste sorting, which can lead to millions of tonnes of construction and demolition waste being redirected annually.

Luke Graham, Head of Research of Pi Labs, said: “With the built world already falling behind on climate targets, we are pleased to share that the findings of our report indicate that AI is set to have a transformative impact on carbon reductions. According to venture funding data from 2023, there is already significant investor interest in AI-driven green solutions aimed at the built environment, however, there is the potential to drive this figure up yet further with a clearer understanding of the positive climate impact and growth potential of these technologies. The good news is that the pace of AI innovation being achieved since the launch of ChatGPT in late 2022 can be put towards significant climate change mitigation by 2030 and as always, the early adopters within the real estate world are set to benefit the most.”

Faisal Butt, Managing Partner and Founder of Pi Labs said: “While Pi Labs has been investing in AI solutions for years, anecdotal evidence indicates generative AI has served as a catalyst for real estate executives to take their wider innovation strategies more seriously. Pi Labs is proud to be revealing, for the first time, the scale and impact that AI can have on the industry and most importantly, in reducing the harmful impact that the built world is having on the environment. Our team of researchers include two former members of the University of Oxford’s Future of Real Estate Initiative. Their knowledge, combined with the insider intel that we have gleaned from some of our start-ups’ success stories to date, has unlocked the capability to produce this groundbreaking report.”

Technology is already playing an important role, with AI revolutionising the way we design, construct, manage, and experience real estate through enhanced efficiency, transparency, and data-driven decision-making. From enhancing design and planning to optimising construction processes, energy management, and corporate decision-making, AI is bringing numerous benefits as it continues to evolve.  

The acceleration of AI in real estate is reliant on access to AI talent and regional differences are already emerging. In 2023, Atomico’s “State of European Tech” report identified that with 108,000 “AI operators”, Europe outgunned the United States (with 87,000). In Sustainably Intelligent, Pi Labs’ analysis of LinkedIn profiles confirmed Europe’s comparative strength in this regard. The research team found approximately 407,429 Europe-based LinkedIn profiles associated with artificial intelligence, compared to approximately 331,000 in North America.

In the interests of balance, the research team were also keen to explore how enhanced AI use within the real estate sector might conversely increase the industry’s carbon footprint. Calculations were undertaken to ascertain the energy consumption and carbon emissions that would result from every real estate professional using generative AI such as ChatGPT-4 for an hour each day.

In the European Union and the UK, this would result in 75,385 kWh of energy demand each day, equivalent to 22.46 tonnes of CO2-e emissions daily. Over a year, this would be the equivalent of 4,099 airline passengers taking a round trip between New York and London. Globally, the figure is 516.6 tonnes of CO2-e emissions daily, or 94,280 airline passengers taking a round trip between New York and London each year.

Launched in 2014, Pi Labs was Europe’s first proptech-focused VC and has been the first institutional investor in some of the most promising AI companies globally that serve the real estate and construction industries including London-based Contilio, Munich-based Conxai, and Tel Aviv-based Firmus.