Each year technology becomes more prevalent in our lives. In many industries it has been changing the way we work for decades. Yet the construction industry seems to be slow to adopt these new technologies. For example, many contractors still use spreadsheets and rulers to perform material take-offs, when there are more accurate electronic ways to get the same information.
A study published in the Journal of Building Engineering in 2021 looked at the current state of the use of artificial intelligence (AI) in the construction industry. It also looked at the challenges and limitations of the technology in this industry.
What does artificial intelligence in construction look like?
Artificial intelligence has several faces in the construction industry. It has the capability to assist in all phases: planning, design, construction, and post-occupancy. Here’s how it’s making its presence known in the industry.
Machine learning uses experience and past data to predict outcomes without needing additional programming. The software has the capacity to notice patterns and “learn” from them, so it knows what to look for next time.
Planning and scheduling
AI applications can select and sequence actions to meet a goal (planning), as well as allocate time and resources to those actions based on the amount available (scheduling).
Robots are being used to construct materials, components, and even whole buildings. The modular construction sector is growing rapidly as the industry becomes more aware of its flexibility in design.
Along with machine learning, AI is allowing software to not only analyze data but make decisions based on the predictable outcomes from a specific action. This allows project teams to make decisions faster and with a greater degree of certainty.
AI also provides an opportunity to make the best decisions with a specific set of constraints.
Challenges in AI Use in Construction
- The construction industry has a history of being slow to adopt technology and even holds the illustrious title of being one of the slowest at incorporating it. There may be several reasons for this reluctance. For one, due to the risky and costly nature of most projects, unproven tech is often not used. Also, since every construction project and site is unique, the value of data collected from one project is limited, unless the AI solution can adapt to a changing environment. And there is a lack of trust in solutions that don’t explain why a certain decision or prediction was made.
- AI systems and data are particularly vulnerable to cyber-attacks. Even small mistakes in data can lead to costly or critical errors on a project that could endanger the life and health of building occupants and construction workers. Also, due to the high cost of projects, contractors and project owners are vulnerable to financial phishing scams and hackers that could manipulate the data for their benefit.
- The much-documented skilled labor shortage in the industry prevents companies from implementing new technologies when they are primarily focused on completing the work in front of them. According to Associated Builders and Contractors, the industry needs to recruit an additional 540,000 workers on top of the normal pace to meet the need in 2023. The more that AI solutions are adopted, there will be an increasing need for AI engineers who have construction experience. This very small population of candidates is going to be hard to find and there will be a great deal of competition in attracting them.
Limitations in AI Use in Construction
- Companies will be limited in their use of AI based on the availability of electricity, computing power, and connectivity. Since some projects take place in areas that are not well connected, it can be difficult to access the software needed to perform these AI functions. These tools require electricity and a connection to the cloud to access shared data and perform calculations. Without these fundamental connections these tools aren’t nearly as powerful. The rise of 4G and 5G communications systems have improved the connectivity issue in many areas.
- There’s no getting around the high cost of purchasing and implementing AI technologies. Robots for the construction industry can cost in the millions of dollars. Software programs with machine learning and automation features may be priced out of reach for smaller companies. Until these tools become more affordable, many companies will be unable to adopt them, even if they would save them money in the long run.
Before AI can be more widely accepted and used in the construction industry, it must address these challenges and limitations. Only time will tell if it has the capability to do so. Meanwhile, small numbers of contractors will continue to test and improve these tools, and reap the benefits of this brave new world we now live in.