When one thinks of artificial intelligence or machine learning, one of the first things that comes to mind is a far-off, sci-fi scenario whereby a man-machine prototype like “The Terminator” takes over the world and all of humanity is at its mercy. That picture of artificial intelligence (AI) is best left in the realm of the imagination, for now at least.
If you’ve ever seen how Netflix suggests movies, used a chatbot for help at your health insurance provider or bank, or used a smartphone, you are already a beneficiary of artificial intelligence. And, if you shop on Amazon or use social media, your search results, feed, and recommendations are based on artificial intelligence. Your online behavior, past purchases, and web searches tailor the results you see, just for you.
Artificial intelligence has its roots in the 1950s when Arthur Samuel, a pioneer in the field, defined it as “the field of study that gives computers the ability to learn without explicitly being programmed.” In other words, computers that learn from experience. Artificial intelligence has also been described simply and concisely as an effort to make machines do things that humans are presently doing better.
Artificial Intelligence and the Construction Industry
Advanced digital technologies, like those above, are right in front of our noses. But the construction industry is known as a slow adopter of technology and one of the least digitized (manual calculations, reports, and practices remain common). Resistance to change to digital experiences and the many manual, repetitive tasks hobble the industry with project delays, cost inefficiencies, and undermined productivity, health, and safety performance.
There is enormous potential to adopt digital technology in the construction industry to improve performance and productivity. While I’m not making any predictions regarding any specifics we may or may not implement here at ConstructConnect, let’s explore a few ways artificial intelligence and the construction industry could make the industry a better place.
Most artificial intelligence people experience today is in the form of machine learning, like when you search Google and see results based on your past online behaviors and interactions. Machine learning aims to replicate human intelligence, but sometimes, machine learning can outperform human insights or decision-making by the sheer volume of its data processing capabilities.
The terms artificial intelligence and machine learning are frequently used interchangeably, but machine learning is a subfield of artificial intelligence.
Construction: An Industry Full of Data, Lots and Lots of Data
Data in the construction industry is greater by volume, more variable and complex, and produced quicker than ever. Data can help make better decisions. Of course, data does no good unless you can use it. But when data, analytics, and human insights combine, it can help improve operational efficiencies, drive growth opportunities, and support wise decision-making. Technology is the bridge between the data and the potential positive outcomes which spring from processing and analyzing the data.
Opportunities and Challenges: Artificial Intelligence in the Construction Industry
Here are six areas where artificial intelligence could benefit the construction industry and their challenges in getting that accomplished. These examples of artificial intelligence have already been successfully deployed in industries like telecommunications and manufacturing to help efficiency, profitability, and safety.
Machine Learning
What it is: Computers designed to think like people to make and execute informed decisions with extensive, variable sets of data.
What it Could Do: This technology has the potential to save costs and reduce waste by making predictions about things like optimizing cuts in steel beams in a building under construction. Amazon used machine learning to find the best package size to ship products safely. The savings were the equivalent of more than 2 billion shipping boxes. Deploying building performance simulations and engineering analyses could enhance construction and design efficiencies using a digital twin of an actual building.
Challenge: Construction industry data is highly dimensional, which means it has many variables like the size and shape of structures or jobsite conditions. The more significant number of variables (i.e., more dimensional) present a challenge to making accurate predictions.
Robotics
What it is: Robots are automated devices that perform physical tasks which resemble human physical activity.
What it Could Do: Robots are already in use in places like welding in automobile manufacturing facilities, replacing knee joints with doctors, and delivering goods to remote locations. The construction industry has seen robots deployed for bricklaying and rebar tying. Robots could get more involved in the construction industry to perform specialized, repetitive tasks or operate in hazardous environments with reduced risks, like on top of tall buildings. With a robot deployed on a jobsite, the engineers, general contractors, and tradespeople could direct their attention to the tasks which demand higher skills.
Challenges: Robots perform best in structured work environments, which poses a challenge on many jobsites where variability frequently occurs in the terrain, operating conditions, and design elements. The high initial cost and maintenance and repair expenses of robotics are also considerable.
Knowledge-Based Systems
What it is: Knowledge-based systems are the field of artificial intelligence where computers make decisions based on existing knowledge.
What it Could Do: Knowledge-based systems can gather and process large amounts of data from various sources and use that data to make complex decisions. The two main parts of a knowledge-based system are the knowledge base that holds the data, and the inference engine, which processes the data in the knowledge base. These tools help store vast quantities of information and produce insights to help people make better decisions. They are even good at demonstrating how they came to the conclusions they reach. Knowledge-based systems are in use by doctors to make more accurate diagnoses.
This artificial intelligence could someday help construction contract management better understand complex issues and improve accuracy. Similar uses of AI could improve health and safety issues by storing data from accidents, then applying those learned experiences to predict risks and provide proactive safety recommendations.
Challenges: Knowledge-based systems can handle vast quantities of data. The challenge with all this data is making sure it is “good,” meaning relevant, accurate, and valuable. Data must be acquired and validated. Since the complexities of the construction industry involve many companies, materials, people, industries, and so forth on a particular project, the data would come from multiple sources and be of differing levels of quality. Proprietary and legal challenges of sharing data are also a hurdle in this area.
Computer Vision
What it is: Equipping and training computers to see and understand images by video, photograph, or in real-time environments.
What it Could Do: Having “an extra set of eyes” with computer vision has seen usage in construction safety applications and progress checks. Collecting, analyzing, and recognizing images with computer vision could be expanded to improve material management capabilities or streamline work procedures.
Challenges: Computer vision successfully deployed needs to understand the entire environment in which it operates, including workers, materials, and actions. That development involves a significant investment in computer vision systems to track and visualize complex and variable construction environments and situations.
Natural Language Processing
What it is: Natural Language Processing, or NLP, enables computer systems to mimic human speech and text capabilities. NLP technologies aim to process text or voice data and comprehend its meaning in the context the writer or speaker intended.
What it Could Do: Voice commands on virtual assistants like Amazon’s Alexa or Apple’s Siri and speech-to-text dictation software have emerged to become a familiar presence in our lives. NLP is also growing in areas to simplify business processes and enhance productivity. Processing unstructured text in construction documents and deriving insights for improved project planning, safety, and material management could benefit trades, contractors, architects, and engineers.
Challenges: With multiple meanings, complex variations, and endless exceptions in human languages, programming language-driven applications are a problematic and intensive task for the emerging field.
Optimization
What it is: Optimization by artificial intelligence is a problem-solving technology that finds the best of all solutions by analyzing and predicting possible outcomes.
What it can do: Optimization aims to increase productivity and efficiency and save time and costs by making choices given limited resources. The healthcare provider, Cardinal Health, uses optimization to find hospital patients at risk and provide personalized, prioritized recommendations for better outcomes. Optimization could provide enhanced work schedules, cut materials costs, or improve energy efficiencies in the construction industry.
Challenges: Optimization needs quite a bit of data to perform predictions and make decisions. Construction variables may include building design, site conditions, material properties, and construction strategies. Real-time processing and large data volume require high-performance computing.
Integrating and Scaling AI Technology
Don’t get boxed in thinking these technologies can only operate isolated and detached. Instead, they have the potential to integrate and grow with the people that operate and use them to improve proficiency, efficiency, and safety. For instance, we may see ways to build better supply chains by leveraging combinations of skilled AI workers, software, robotics, and predictive technology powered by machine learning and artificial intelligence. Building Information Modeling (BIM) has already been used with AI to some degree and with mixed results, leaving this an area ripe for further exploration and development.
Building Information Modeling, the state-of-the-art system in the construction industry, may further engage with artificial intelligence to improve reliability and accuracy when estimating time and costs. A system that could help owners, architects, engineers, the trades, maintenance, and support staff with detail-rich automated systems to design, check for code compliance, and diagnose problems before they happen. Robotic systems could project BIM images and installation guides right onto a workspace like a floor or a wall or be used to reveal hidden objects behind walls for plumbing or HVAC maintenance.
Stay Tuned for Developments in the Field of Artificial Intelligence
We are beneficiaries of the direction and pace of technological advances to the degree technology improves our lives and environment. The construction industry stands ready to receive the potential process, health and safety, and production benefits from developing and implementing artificial intelligence. One cannot be sure where this technology will go, but robots taking over the world remains safely planted in the domain of authors and screenwriters.
Future systems may allow teams to collaborate and improve performance by capturing data and deploying a single, unified, shared data stream during the entire life cycle of a construction project, from ideation and preconstruction to construction and maintenance.
It is an interesting and exciting time to consider the impressive potential of artificial intelligence and witness the developments in the field. Stay in touch with the changes and stay tuned for what artificial intelligence may do for the construction industry.
Doug Dockery is the Chief Technology Officer for ConstructConnect, overseeing IT, application development, DevOps, QA, data management, and our PMO. Prior to joining ConstructConnect, Doug served as a global technology and digital transformation leader for LeadingAgile, CentralSquare Technologies, CA Technologies, Rally Software, and a number of other organizations. A native Floridian, Doug resides in the Orlando, Florida area with his family. He is also active within the Central Florida community by serving a number of area nonprofits.
Responses