Written by Peder Aaby, Chief Technology Officer in Sparkel. Msc. Cybernetics and Robots with specialization in machine learning

Click here to see Sparkel.io’s offering for GPT4 based BIM and PDF quantity takeoffs

Building Information Modelling (BIM) was supposed to take the global digitalization trend to construction and be the true enabler for productivity growth.

It never happened.

"Globally, construction sector labor-productivity growth averaged 1 percent a year over the past two decades, compared with 2.8 percent for the total world economy and 3.6 percent for manufacturing."

- REINVENTING CONSTRUCTION: A ROUTE TO HIGHER PRODUCTIVITY, McKinsey

While BIM has numerous benefits, the costs associated with having having an up-to-date information model are very high. And these models and the platforms to interact with them are often heavy and technical, which never really make the benefits apparent or useful for the regular construction worker.

That said, the recent development in AI and specifically large language models will be a revolution for the construction industry and BIM usage. Suddenly we can now abstract the technical nature of these models and interact with them with natural language. And AI can automatically fill in and correct the information gaps in existing models. Both making the models 10x more accessible and give a very large cost decrease in creating and maintaining a up-to-date building information model.

The following post will go into the basics of this AI revolution and how it deeply affect how we think about buildings and building information models.

Brief Explanation of AI

AI, or artificial intelligence, refers to the simulation of human intelligence in machines that are programmed to think and learn like humans. AI technologies enable machines to perform tasks that would normally require human intelligence, such as recognizing speech, making decisions, and identifying data patterns. These technologies are advancing exponentially and are becoming integrated in a wide range of industries, from healthcare and finance to transportation and manufacturing. And the recent advances in large language models is nothing less than a revolution.

Brief Explanation of BIM

In order to comprehend the rest of the article, a quick refresher of the BIM concept is useful. Building Information Modeling (BIM) is a digital representation of a building’s physical and functional characteristics. It is used by architects, engineers, and construction professionals to design, construct, and operate a building. Modeling software is used to create a 3D model of the building, which includes information about the materials, structure, and systems used in its construction. The importance of BIM lies in its ability to provide a comprehensive view of a building’s design and construction process. BIM allows project teams to collaborate more effectively, identify potential issues before construction begins, and reduce errors and costs.

Challenges with BIM

Despite the many benefits that BIM offers, there are also challenges. One criticism is that BIM can be difficult and expensive to implement. Implementing BIM requires significant changes to traditional workflows, as well as new hardware and software investments. Another criticism of BIM is that it can be overly complex and difficult to use, especially for smaller construction firms with limited resources. Additionally, there are concerns about the lack of interoperability between different BIM software programs, which can make collaboration between different stakeholders more difficult.

Can AI Solve BIM Challenges?

While AI cannot completely solve all the criticisms of BIM, it can address many of the challenges associated with the technology. For example, AI can help to streamline the implementation process by automating certain tasks, such as material takeoffs, and reducing the need for manual data entry. AI can also make BIM more user-friendly by simplifying complex workflows and providing more intuitive interfaces. Ie. simply querying information or quantities from your BIM model in natural language. Additionally, AI can help to address interoperability issues by providing a common platform for different stakeholders to collaborate on a project. While AI cannot replace human expertise in the construction process, it can help to augment it by providing data-driven insights and reducing the risk of human error. Therefore, while AI cannot solve all the criticisms of BIM, will help to make the technology more accessible, efficient, and easier to use.

Benefits of AI in BIM

The benefits of AI in BIM are numerous. First and foremost, AI can and will save time and reduce errors. With AI, contractors can extract material quantities by just asking for it, reducing the time and effort required for manual calculations. This can also reduce errors caused by manual calculations, ensuring accuracy in the construction process. However, the latest rapid advances in AI has been in the field of large language models. These models cannot consume BIM data directly. Therefore, having an interface between AI and the BIM model is a necessity. Meaning, AI needs a platform like Sparkel to function for BIM. The platform at hand works as a layer between large language models and BIM.

Use Cases of AI in BIM

There are several use cases of AI in BIM. One of the most common use cases is in the estimation of material quantities. AI algorithms can analyse 3D models and extract quantities of materials automatically, saving time and increasing accuracy.

Another use case is in the scheduling of construction projects. AI algorithms can analyse 3D models and predict the time required to complete a project based on past data. This can help contractors to schedule projects more accurately and reduce delays.

AI can also be used to analyse construction designs and identify potential issues before construction begins. For instance, AI algorithms can identify potential clashes between different building components, reducing the need for rework during the construction process.

Text to Quantity Extractions

Text to quantity extraction is a method of extracting material quantities from a BIM model by analyzing the model’s metadata. Traditionally, this process required manual filtering of the model, searching for properties, and setting up the material extraction. This process is often time-consuming and prone to errors when done manually.