With nearly 1 in 4 US construction workers over the age of 55 and 1.9 million among the workforce expected to leave the industry this year, the size and scope of the construction labor problem is taking on new proportions. The current shortage averages around 400,000 workers per month, particularly in the skilled trades. The Associated Builders and Contractors Association (ABC) estimates that the US construction industry must actually recruit closer to half a million new workers to meet demand in 2024 alone, and that’s if construction starts slow dramatically heading into 2025.(1) None of this takes into account the anticipated growth in mega-projects and the jobs that will be created due to initiatives like the Bipartisan Infrastructure Law (BIL), particularly in the areas of infrastructure and energy. The BIL alone could expand the labor gap by an additional 345,000 jobs by 2026-27.(2)
To say that the construction industry has a labor “gap” is an understatement. The demand for construction labor is a swiftly expanding chasm. Increases in trade school enrollment, training and recruitment won’t make a dent in it. Transformational change across every aspect of how we do business will be required to bridge it.
The labor shortage is an integral part of ongoing stagnation in construction productivity and growth. A recent study from the Becker Friedman Institute of Economics at the University of Chicago maintains that not only has industry productivity stagnated, it has declined in the past 5 decades, with a 40% decrease in productivity since 2020. Meanwhile the rest of the economy over this period has doubled.(3) The productivity landslide will only be turned around by more rapidly advancing process innovation, digitization, and automation.
Technology has long been acknowledged as the key that will unlock productivity and growth, just as it has done for literally every other major global industry. Digital transformation in construction holding the promise of increasing productivity by 15 percent, and the use of real time data can help construction work be 50% more productive.(4) The question is, with so much investment in construction technology over the past decade, why haven’t we made greater inroads in solving both the labor and productivity problems?
The industry is awash platforms and point solutions aimed at doing just that, but we’re still accessing only a small fraction of the data each project creates, and we still haven’t turned the tide on productivity. Technology is something we’ve largely force fit into a very traditional process, rather than using it to rebuild the process from the ground up.
Artificial Intelligence (AI) is the change agent. It has the potential to be cut through the noise of construction technology and the data it has created by collecting vast data sets, locating meaningful connections and intelligence, and then putting this intelligence to improve processes, automate redundant work, streamline and de-risk project planning and controls. AI will help unleash the potential of technology that is already in place, layering in human experience to the equation to truly unlock construction productivity and shore up the labor gap.
Where will we get all the people we need? Gen Z was recently dubbed “the toolbelt generation,”(5) because they’re entering trade schools and apprenticeships in outsized numbers to their predecessors. We must bear in mind that they are also the first digitally native generation. They don’t need to be ‘upskilled’ like the current workforce, and they’re already using ChatGPT on a daily basis. As they enter the construction workforce, digitization and the use of technology can help attract them to the field, and their tech-fluency can make them an essential part of the construction transformation.
This is important, because technology alone won’t make a bit of difference if it isn’t broadly deployed across the construction process, from development and design to preconstruction and construction, and it won’t have real impact unless it is broadly adopted by stakeholders across the value chain, from project executives to skilled labor to boots on the ground. Before digging into how AI can help, let’s first acknowledge that the goal must be to find and address pain points and process inefficiency, not install new IT solutions.
In our work at Slate, these are some of the client use cases that have shown to have the biggest potential to impact the labor shortage.
AI-enhanced computational design can reduce human error and rework. It’s widely understood that digitized design & BIM allow for greater accuracy and faster, less impactful trial and error. AI can be used to identify conflict and potential issues, and fix them before they become problems in the field. This saves time and labor.
Automation for direct labor savings is table stakes. We need to look for more nuanced, impactful uses of Generative AI. AI can be used to automate repetitive tasks and reduce human error that naturally occur with things like data entry. That’s table stakes. To really impact project productivity from the start, we need better takeoff and estimating and better schedules. Slate is working with a Tier 1 client who is using AI-driven automation to productize design and accelerate preconstruction. The primary ROI for them isn’t just the efficiency they gain in the design stage, automation helps them build more, and be more profitable.
The Slate system ingests data inputs, design attributes and characteristics generating the BIM model and tagging it to disparate data to surfaces a single-source-of truth that can be customized to interact with multi-disciplinary contributors based on their role in the process. This gives individuals and teams unprecedented data and document availability and creates a platform for productive collaboration. All design, estimating and manufacturing documentation is automated, and changes that take place across the process are captured and changes to documentation are automatically generated. What took days and even weeks, now takes minutes when using custom digital designs that the client defines.
Better, faster design makes for smoother project deployment, and it sets the stage for what we believe is one of the primary levers AI can pull in the effort to bridge the labor gap: the reduction of rework.
Do it right the first time. Eliminate rework
Thirty percent of construction work is rework. If we reduce rework, we shore up one of the biggest drains on labor resources. Rework is one of the primary drivers of missed deadlines, schedule overruns and budget blowouts. It has a compounding impact on labor resources: more people are required to fix a problem than to do it right the first time. Imagine the construction labor that could be reduced or better yet, redeployed on a new project, if we could reduce, or even eliminate rework.
Here are some of the ways AI can help construction do things right the first time:
Better Data means less rework.
AI has the ability to overcome two of the biggest contributors to rework, lack of data and poor communication. In fact, 52% of global construction rework is the result of these two failure points. bad data and lack of communication. At the core of the AI engine we have built at Slate Technologies, is the ability to collect, analyze and contextually connect data, regardless of source, so that it can be surfaced to teams no matter whether they’re onsite or in an executive suite. Using Conversational AI, this data can be queried, documents and information across multiple platforms can be quickly located. Generative AI can automate reporting. Predictive AI can be used for better procurement, schedule accuracy and automation, and project management. With AI, projects can proceed faster and without as many mistakes.
Plan projects and labor better. Do more with less. AI-driven scheduling.
It won’t be enough to recruit and train new workers, the industry must learn how to do more, and do better, with less, or rather with fewer workers. The key to this, and the backbone of every project is its schedule. Schedules make or break projects, and even with the best tools available, they are only as good as their human inputs and only if they are accurately maintained.
AI-tools abound in this space, most of them solving part of the problem by providing for optioneering, flagging issues, or surfacing single-project intelligence. The gold standard will be when an AI-engine can not only help design a good schedule, but also manage change over time to keep a schedule accurate in real time, and then feed these learnings back into the next schedule, and the one after that with insights such as, “Supplier Y has an average delay of 65 days on the past 30 projects,” or “The average float on the past 30 projects was 30% lower than actuals.”
At Slate we’re currently engaged in supporting a client in creating an AI-enabled, comprehensive system and data framework to connect diverse processes and schedules into a single, unified system capable of actively managing, predicting delays, and optimizing schedules for design, construction, manufacturing, and procurement.
Use AI to manage project controls better.
Let’s face it, managing a project and delivering it on time is a herculean task. The typical large construction project generates 30 million documents, and even the 8000 or so individual contributors to the typical project can’t realistically be expected to get their arms around project data of that magnitude, much less communicate it out in real time. The typical construction employee, regardless of role, wastes around 30% of their time trying to get their hands on information and documents.
Slate uses AI not only to aggregate data, we put it to work helping teams utilize their data to run quality management and progress tracking and controls directly on the BIM and to run digital progress reporting. These applications of AI are saving labor resources, and they’re also helping avoid rework. These applications can be leveraged to manage other project controls like safety.
Data-driven Insights and Recommendations. Turning lessons learned into better decision making, more accurate labor planning, more nimble project pipelines and bigger profits.
Every project team starts with the intention of capturing lessons learned and feeding it back into organizational intelligence to make the next project better, and a lot of teams do conduct a formal review and analysis. Do those lessons learned actually feed continual improvement? Not often. This is likely the biggest impact AI will have on construction – the ability to capture, contextualize lessons across every aspect of a project pipeline, revealing patterns and trends that help inform more efficient allocation of existing labor resources, better preconstruction, planning, project management and control.
And, so you ask, will AI take construction jobs?
While it is true that robots and “cobots” can replace and/or enhance workers performance in the field, and it is undeniable that technology will be at the heart of how we solve the labor shortage, the future rest squarely on the shoulders of humans, their experience and intelligence, and their ability to use AI as a tool.
AI will be used to augment human experience and supercharge the performance of individuals and teams. It will help us connect the dots in our data and in doing so it will help us communicate better, make rework a problem of the past, and force us to reconsider how we use labor to begin with. Ultimately it is already helping those who use it deliver better projects, faster and more profitably, giving innovators a rapidly accelerating competitive edge.
“AI won’t replace humans, but humans with AI will replace those without.” Karim R. Lakhani, Harvard Business School
Discussion
Be the first to leave a comment.
You must be a member of the BuiltWorlds community to join the discussion.