Artificial intelligence is the talk of the town in every industry these days, and construction is no exception. Digitalization is imperative to improve operational efficiency in the Architecture, Engineering, and Construction (AEC) industry in the future, and AI is one of the most impactful technologies to make that happen. Here’s how.
The construction industry is well known to be one of the least digitalized industries in the world, with almost stagnant productivity growth over the past two decades – growing at only 1%.
As opposed to more traditional IT, which is more geared toward rule-based, predictable processes as we see in manufacturing. AI has the potential to deliver dynamic digitalization at scale where operations are similar, but not identical, as is the case in construction.
Turner Construction, the largest contractor in the US, predicts that artificial intelligence will transform the industry in the next 10 years more than any technology in the past 100 years.
This may instinctively be perceived as an- overly techno-centric and audacious statement, but if we dive into the range of potential use cases for AI and algorithmic design tools as a broad term in construction, this may be closer to reality than our initial gut reflex might think.
Just like the introduction of the internet gave birth to online banking, and the launch of the smartphone brought technology into our everyday lives in a manner we might never have imagined in advance, AI has the transformative power to usher in an age of digital transformation in the AEC industry.
Looking at the entire AEC value chain, AI has a part to play in every step of the way from design throughout construction all the way to building maintenance and eventual decommissioning.
Designers and architects may be some of the early adopters of advanced data modeling tools in the industry, having used model-driven design for decades already. With solutions like Spacemaker – now Autodesk Forma, Spacio, Arealize, and Parallelo, just to name a few coming out of the Norwegian contech startup scene, designers can generate a wide range of optimized building design suggestions in seconds. Further on, tools like Consigli may evaluate whether those designs are up to speed with current technical standards and regulations.
To take this even further, technology can also help in sourcing the right products and materials for any project – a process that is tedious and time-consuming. A fine-tuned language model like we see in ChatGPT can cut research time, suggesting the best product matches and even answering questions relating to product performance.
The use of AI for increased bidding intelligence may increase the value of historical data, as contracts will be able to analyze vast amounts of data to both more precisely predict the probability of certain bids, but also heighten the quality of tenders by using AI to check for potential inconsistencies.
However, where things really get exciting is when AI is applied to the construction phase. This is where the virtual world of data meets the hard physical reality of a construction project. The amount of data produced at any given construction project is staggering, but this data is most often entered in retrospect ad used for reporting purposes.
By capturing and processing data in near real-time, solutions like Ditio will be able to present those data as actionable insights to site managers, thus directly impacting project performance on a granular level.
At Skanska, we are currently using those data sets to automate and optimize mass hauling. Thus, reducing fuel usage and lowering CO2 emissions. Initial research found that machines on an average civil construction site are idling about half the time. Much of this time is when haulers are standing in line or waiting on another machine to complete its task. To reduce this time, the underlying algorithms enable “live, dynamic routing” that reorganizes circulation routes and staging patterns on-site to minimize idling. This solution, which will be available for the entire industry through Ditio, will be able to suggest optimal routing and machine fleet composition day by day, as environmental conditions change.
Based on the same data sets, a new set of machine learning algorithms may be utilized for predictive maintenance of equipment and vehicles as well. Such a system would take haulers specific duties and cargo into account. For example, trucks that are used to haul boulders will need checkups more often than ones used to haul soil. By being able to predict more precisely when a machine will break down using both machine and project data, it is possible to avoid a halt in production due to unexpected machine failures, as well as reduce unnecessary repairs.
AI can also be used to identify potential safety hazards before they occur, improving job site safety This may be everything from using image recognition on drone footage to predict potential safety hazards to check for proper use of PPE (personal protective equipment) by workers on site.
Those same images me be utilized to automatically track project progress and performance, as well as to check for structural integrity in concrete constructions are just some examples of how data may be leveraged to increase operational performance in the construction phase.
And that is without even touching upon the topic of robotics, exemplified through Spot from Boston Dynamics and Siteprint from HP. While it is the robot’s physical presence that impresses us, it is the software and internal algorithms that enable their performance. But that’s a topic that deserves a blog post on its own.
Although there are a multitude of promising use cases, getting started with AI in the construction industry is not something that is done overnight. The IT landscape in the industry tends to be fragmented and is characterized by data stuck in siloed applications. A long-term commitment to a consistent data platform that maintains a single source of truth is imperative to establish a common ground of trustworthy and unambiguous data. This will require investments in both technology and competence.
There has been, and still is a cultural barrier to technology adoption in the industry. According to Constructiondive, there are concerns related to AI’s potential to eliminate jobs. A 2023 survey by Pew Research found that 32% of U.S. workers think AI will hurt the workforce more than help, while just 13% believe the reverse.
However, it is important to note that construction is an industry where the workforce has been shrinking, and AI may prove to be a necessary ingredient to close the gap and let contractors handle more work with a decreasing talent pool.
At the end of the day, using AI should never be a goal by itself, but a means to provide an algorithmic approach to a dynamic environment. To reap the potential benefits of AI in construction, it is crucial to combine technological insight with deep industry knowledge to identify high-impact use cases, thus avoiding chasing solutions looking for problems.