Five Reasons to Use Generative AI to Automate Building Designs


The construction industry has dabbled in artificial intelligence (AI) for routine tasks like scheduling and document analysis. But generative AI is a game changer, says Francesco Iorio, CEO of Augmenta – one that promises to transform the way buildings are designed – lowering costs, increasing productivity, and reducing waste. 

Tools like ChatGPT and DALL-E use large-scale machine learning (ML) models and access large amounts of labeled and meaningful data to provide insightful responses to queries in text and images. But some industries have limited access to data sets to train ML models, making it difficult to reap the benefits of using generative AI to solve real-world problems.

The construction industry is a case in point. There is no single repository containing labeled data of engineering drawings for buildings. That’s because engineering firms are secretive with their data and are not inclined to share their intellectual property. One of the consequences is that      the construction industry has been held back by antiquated design methods. Existing legacy tools used to design buildings and their systems are little better than electronic pencil on paper, leading to unbuildable designs, lack of coordination among trades, and wasted time and materials when work must inevitably be redone. 

That said, even state-of-the-art generative AI models such as ChatGPT – which draw on very large, varied, and detailed datasets to train complex models – can produce erroneous results while exhibiting full confidence in the outputs. In the case of ChatGPT, the consequences of making a mistake are relatively low. But the stakes are simply too high in engineering, which requires more than large black-box mathematical models to adopt generative AI safely and effectively. 

A Paradigm Shift in Building Design

Thankfully, there is a novel, hybrid approach to rule-based AI systems that can produce new and valid data in the form of generated designs that can be used to train ML models. The most valuable application of this approach is to automate building design. Not only does it accelerate the end-to-end design process from months to days, but it also provides an unprecedented level of insight for developers, architects, and engineers to help them make more informed decisions related to cost, schedules, and efficiency. 

See More: The Risks & Rewards of Generative AI

Benefits of Automated Building Design in Construction

Let’s take a closer look at what the construction industry can achieve by automating the building design process. 

1. Design high-performing, code-compliant buildings – reducing risk, delays, and rework

Today, architects and consulting engineers who create high-level building designs don’t have the time or adequate information required to develop constructible systems. For example, the design process for mechanical, electrical,    plumbing and structural systems (MEP/S) is extremely complex, time-consuming, and error-prone. It is also one of the leading causes of errors, delay, risk, and uncertainty. 

By automating design, the rate of the design and construction process can be dramatically accelerated, reducing construction schedules by months to create functional buildings for residential and commercial purposes faster. And by lowering risk and uncertainty and eliminating rework that can add an average of six percentOpens a new window in costs, developers can better plan and budget for their projects while contractors can bid on jobs more accurately.

2. Build a better, more sustainable world

The construction industry is a prime consumer of energy and materials. According to the Digital Transition and Waste Management in Architecture, Engineering, Construction, and Operations Industry report Opens a new window by Frontiers in Energy Research, up to 30 percent of new building materials go to waste due to design errors and rework. These errors can be all but eliminated using an automated design system. 

Generative AI also enables multiple design alternatives to be created in parallel, helping to find ways to develop buildings that perform better using less material. It can also help to drive energy efficiency – a critical capability considering the UN Environment Program’s findings that buildings consume about 40 percent of global energy and resources. Now developers can understand their choices: to optimize solely for cost and schedule or to also design for more sustainable material use and operation. Automatically producing designs at a high level of detail ensures they only order what they need, reducing material waste.

3. Improved efficiency and cost-effectiveness

By leveraging generative AI to optimize the design and performance of buildings, not only can the industry reduce its carbon footprint – it can also do it more productively and cost-effectively. A few years ago, it was estimated that architects, engineers, and construction (AEC) professionals spent approximately 20 percent of their time resolving mistakes and conflicts arising from design and coordination errors. On a global scale, this equated to $280 billion in reworkOpens a new window . These numbers have undoubtedly gone up as talent scarcity and demand for new construction intensify (which we’ll cover below). 

Generative AI brings a level of automation to the design and construction process that enables AEC professionals to create optimal designs in hours instead of weeks and dramatically reduces construction errors. And because designs are created with high certainty, design professionals can be more productive, spending less time on rework and errors.

4. Ease the pressure of talent scarcity

The construction industry faces a massive shortage of trained and experienced people to meet the current pipeline of greenlit projects. In fact, some of the industry’s largest unions are predicting a gap in skilled tradespeople in the United States. 

Automating design means enabling individuals at architecture firms, engineering firms and contractors to have the experience of a seasoned foreman at their fingertips, allowing even junior designers and engineers to create designs that are constructible and code compliant. It also frees those same individuals from the traditional and tedious approach to design work. Instead, they can spend their time truly understanding the needs of their customers, exploring design options and trade-offs to arrive at the best possible design.

5. Scale to meet the demand for new construction 

According to a 2023 engineering and construction industry outlook report by DeloitteOpens a new window , there’s no shortage of investment in new construction projects. In the United States, new housing units are expected to reach 1.55 million annuallyOpens a new window – up from 583,000 in 2009. Given the shortage of skilled workers coupled with high turnover, the industry must learn how to do more with less. 

Generative AI promises to deliver scale across the entire construction ecosystem. Contractors can scale their design capabilities without being limited by the availability of talent or retention. Part suppliers can automate and scale their prefabrication services, including the ability to sell entire, purpose-designed assemblies rather than just parts. The construction industry can finally play catch-up by embracing this emerging technology.

A Solid Foundation to Build On 

There’s no doubt that the construction industry is ripe for disruption. Generative AI has the potential to radically change the course of construction history – fundamentally transforming the way we design buildings and the design of those buildings themselves. And while AI has already delivered some results in the field, it’s clear that the best is yet to come. 

What are your thoughts on generative AI enabling automation? Share with us on FacebookOpens a new window , TwitterOpens a new window , and LinkedInOpens a new window . We’d love to hear from you!

Image Source: Shutterstock