globally, individuals and businesses spend more than $10 trillion annually onconstruction-related activities , with growth expected to continue at 4.2% through 2023. part of this massive spending is being driven and enabled by rapidly changing technological advances that are impacting all areas of the ecosystem.in its2020 report,Construction's Next Normal: How Disruptive Innovation is Shaping the World's Largest Ecosystem, McKinsey identified a growing interest in solutions that incorporate artificial intelligence (AI).
aI in construction has the potential to create value across the project lifecycle: design, bidding and financing, procurement and construction, operations and asset management, and business model innovation. in the construction industry, AI can help overcome some of the toughest challenges across the industry, including safety issues, labor shortages, costs, and delays.
with increasingly lower barriers to market entry and accelerating advances in AI, machine learning (ML), and analytics, we expect AI (and the allocation of resources to AI) to play an increasingly important role in the construction industry over the next few years.
in this article, you'll learn how AI is used in construction and the top 10 benefits of using AI in construction.
what do we mean by artificial intelligence and machine learning in construction?
artificial intelligence (AI) is an umbrella term that describes when machines mimic human cognitive functions, such as problem solving, pattern recognition, and learning. machine learning is a subset of AI. machine learning is a branch of artificial intelligence that uses statistical techniques to give computer systems the ability to "learn" from data without being programmed. the more data a machine is exposed to, the better it understands and provides insights.
bob Banfield, a machine learning engineer at Trimble, explained deep learning in the construction space
"Machine learning involves a lot of algorithms. as a simple example, if you're trying to figure out whether you're likely to get a certain disease, one type of learning algorithm might work through a decision tree of questions, like, 'How old are you?' Then, 'Okay, do you exercise?' if you answer yes, you go down one branch, if you answer no, you go down another branch, and so on. That's a perfectly valid machine learning algorithm. it 's like the game of Twenty Questions that you might have played as a kid, except that in machine learning, these questions are automatically generated."
when applied to construction, the "questions" and algorithms become much more complex. for example, a machine learning program can track and evaluate the progress of a grading plan to identify schedule risks early. the algorithm can "ask questions" about cut and fill measurements, machine uptime and downtime, weather patterns, previous projects, or any number of inputs to generate a risk score and determine if alerts are needed.
aI and machine learning for smart construction
the potential for machine learning and AI in construction is enormous. requests for information, issue reports, and order changes are commonplace in the industry. machine learning is like a smart assistant that can scrutinize this reams of data, and then alert project managers to important things that need attention. many applications are already using AI in this way. they range from routine spam email filtering to advanced safety monitoring.
10 Examples of AI in Construction
1. avoid cost overruns
most large projects go over budget despite hiring the best project teams. Artificial neural networks are used in projects to predict cost overruns based on factors such as project size, contract type, and the competency level of the project manager. predictive models use historical data, such as planned start and end dates, to envision a realistic timeline for future projects. AI helps employees remotely access real-world training materials to quickly improve their skills and knowledge, which reduces the time it takes to onboard new human resources to a project. This results in faster project delivery.
2. aI for better building design with generative design
building information modeling (BIM) is a 3D model-based process that provides architecture, engineering, and construction professionals with insights to efficiently plan, design, construct, and manage buildings and infrastructure. to plan and design the construction of a project, the 3D model must take into account the architectural, engineering, mechanical, electrical, and plumbing (MEP) plans and the sequence of activities of each team. The challenge is to ensure that the different models of the sub-teams do not conflict with each other.
the industry is leveraging machine learning in the form of AI-powered generative design to identify and mitigate conflicts between different models created by different teams to avoid rework. there is software that uses machine learning algorithms to explore all variations of a solution and generate design alternatives. users set requirements for the model, and generative design software creates a 3D model optimized for the constraints, learning through iteration until the ideal model emerges.
3. risk mitigation
every construction project has many forms of risk, including quality, safety, time, and cost risks. the larger the project, the greater the risk, as multiple contractors are working on the site simultaneously. today, there are AI and machine learning solutions that general contractors use to monitor and prioritize risks on the job site, allowing project teams to focus their limited time and resources on the biggest risks. AI is used to automatically prioritize issues. subcontractors are evaluated based on their risk score, so construction managers can work closely with high-risk teams to mitigate risks.
4. project planning
a construction intelligence company launched in 2017 with the promise that robots and artificial intelligence hold the key to solving construction projects that are behind schedule or over budget. The company uses robots to autonomously capture 3D scans of construction sites, then feeds that data into a deep neural network to categorize how far along different subprojects are. if work appears to be off plan, a management team can step in and deal with small issues before they become big problems. future algorithms will use an AI technique called "reinforcement learning." This technique allows algorithms to learn based on trial and error. it can evaluate infinite combinations and alternatives based on similar projects. it will optimize the best path and correct itself over time, which will help with project planning.
5. aI to boost productivity on the shop floor
companies are popping up that offer self-driving construction machines to perform repetitive tasks like pouring concrete, bricklaying, welding, and demolition more efficiently than humans. excavation and preparation tasks are performed by autonomous or semi-autonomous bulldozers, which can prepare the job site to exact specifications with the help of human programmers. This frees up time for human construction workers and reduces the overall time required to complete a project. Project managers can also track work on the job site in real time. project managers use facial recognition, on-site cameras, and similar technologies to assess workers' productivity and adherence to procedures.
6. aI for construction safety
construction workers are five times more likely to die on the job than other workers. according to the U.S. Occupational Safety and Health Administration (OSHA), falls are the leading cause of private sector fatalities (excluding highway crashes) in construction, followed by being struck by objects, electrocution, and being pinched/trapped. one Boston-based construction technology company has developed an algorithm that analyzes photos of job sites to scan for safety hazards, such as workers not wearing protective gear, and correlates the images with accident records. The company says it could potentially calculate a risk rating for a project and hold safety briefings if a high threat is detected. in 2020, it even began scoring and publishing safety scores for each U.S. state based on its compliance with COVID-19 regulations.
7. AI will solve labor shortages
labor shortages and a desire to increase the industry's low productivity are driving construction companies to invest in AI and data science. a 2017 McKinsey report found that construction companies can increase productivity by up to 50% through real-time data analytics. Construction companies are beginning to use AI and machine learning to better plan the deployment of labor and machines across jobs.
with robots that continuously assess the progress of work and the location of workers and equipment, project managers can instantly see which job sites have enough workers and equipment to complete the project on schedule and which sites need to be staffed with additional labor.
aI-powered robots like Spot the Dog can autonomously scan job sites every night to monitor progress, allowing large contractors like Mortenson to complete more work in areas where skilled labor is scarce.
8. offsite construction
construction companies are increasingly relying on offsite factories, where automated robots assemble components of a building, which are then assembled by humans on site. structures such as walls can be completed in an assembly-line fashion by autonomous machines more efficiently than humans, and details such as plumbing, HVAC, and electrical systems can be finished by humans as the structures fit together.
9. aI and big data in construction
in an era where vast amounts of data are generated every day, AI systems are exposed to an infinite amount of data to learn and improve every day. every shop floor is a potential source of data for AI. data generated from images captured by mobile devices, drone video, security sensors, building information modeling (BIM), and more has become a pool of information, which presents an opportunity for construction industry professionals and customers to analyze and leverage insights generated from data with the help of AI and machine learning systems.
10. aI for post-construction
building managers can use AI even after construction is complete. by gathering information about structures through sensors, drones, and other wireless technologies, advanced analytics and AI-powered algorithms can gain valuable insights into the operation and performance of buildings, bridges, roads, and virtually anything else in the built environment. This means AI can be used to monitor for problems, determine when proactive maintenance is needed, or direct human behavior for optimal security and safety.
the future of AI in the construction industry
robotics, AI, and the Internet of Things can reduce construction costs by up to 20%. engineers can wear VR goggles and send a mini robot into a building under construction. the robot uses a camera to track the progress of the work. AI is being used to plan the routes of electrical and plumbing systems in modern buildings. companies are using AI to develop workplace safety systems. AI is being used to track the real-time interactions of workers, machines, and objects on the job site and alert supervisors to potential safety issues, construction errors, and productivity problems.
despite predictions of massive job losses, AI is unlikely to replace the human workforce. instead, it will change the business model of the construction industry, reduce costly errors, reduce workplace injuries, and make building operations more efficient.
leaders of construction companies should prioritize their investments based on where AI can have the greatest impact on their company's unique needs. Early movers can set the direction of the industry and benefit in the short and long term.
