Much is being written in the media about how Artificial Intelligence (AI) is transforming the workplace. Almost every software vendor, and even some hardware vendors, are pitching the incorporation of AI in aspects of their technology. This is no different in the construction industry, with new innovations entering the marketplace almost every week it seems. However, when looking past the hype regarding what AI can do for the industry, what actually is it doing? How are contractors actually using AI today, not just the promise of AI for tomorrow. This article provides examples of how contractors are using AI in real-world implementations within the industry. This is by no means an exhaustive list. The intent is to show that AI is being used in construction now. While some of those leveraging AI are large general contractors with extensive IT staff, not all are.
Winning Work and Pre-Construction
Before any construction can begin, a contractor first has to win the work. Several contractors are using AI to improve their win or hit rate by taking historical customer relationship manager (CRM) and enterprise resource planning (ERP) data to identify which opportunities have the highest probability of being won. By analyzing past data, contractors are able to determine that based on certain demographic data for the project, which team combination (both internal and external), they are most likely to win with when competing against certain other contractors. Essentially augmenting the Go/No Go decision.
Of course, having a good estimate is a key component to winning a project too. AI is being used by contractors to assist in this work effort as well. Several companies are experimenting with AI assisted takeoff to speed up the take off process. While the takeoff still needs to be reviewed, contractors using this approach have found significant time savings. In addition to AI assisted takeoff, AI is being used to pull data from specification documents (in addition to drawings) allowing contractors to better capture project requirements that impact the estimate. By combining this with historical cost data, as well as change orders on similar projects, to identify risks associated with similar work and details on subcontractor performance and material buying history, AI is being used to develop a better estimate. This helps in reducing the risk of missing critical information for the project. Additionally, AI utilizes historical data to predict material pricing changes, potential change order impact and subcontractor performance. These factors significantly influence the project’s risk and budget.
For those leveraging building information modeling (BIM) on projects, there are certainly plenty of articles about contractors leveraging AI against a digital twin of a building to identify potential risks and optimal schedule/resource allocation before construction even begins, however, this takes considerable resources to accomplish and is it not practical for all projects.
Performing the Work
There are a number of areas in which AI is being used to assist in the performance of work on a construction job. Several contractors are using AI assisted scheduling. By analyzing data from past projects, they are using AI to identify which tasks for a similar project are likely to take longer or get delayed. Using this information, they are able to better determine which types of additional resources are needed, where to apply them and when to minimize delays. Similarly, one contractor successfully experimented using IoT devices on jobsite equipment (i.e., cranes) and tags on materials and supplies, to leverage IT to assist with logistics planning, identifying potential bottlenecks, and speed the time locating materials and identifying potential safety issues based on equipment usage.
AI is being used to assist in excavation and prep work on jobsites. Contractors are using autonomous and semi-autonomous equipment combined with AI to excavate and prepare jobsites with reduced human interaction, reducing the overall time required to complete the project. By improved consistency and tracking, the automated equipment also reduces the amount of re-work that needs to be done. AI is being used to help maintain the equipment too. Several contractors are using AI against historical equipment management and maintenance data, to identify when equipment is most likely to fail so that maintenance work can be performed proactively minimizing delays, allowing equipment to run longer and reducing the additional costs of rental equipment while company owned equipment is out of commission.
Making sure you have the right team mix on a project can be challenging. You don’t want to have too many inexperienced people working together. One enterprising contractor is using a combination of training, certification and employment data from HR, and historical project data from job cost/project management systems and safety, to determine the optimal project team mix for projects. They look at a combination of years of experience, skills and certifications, training provided to the team, subcontractor performance and how safety conscious the individual team members are. They make sure that a team has the appropriate mix of a younger team (with better technical capabilities and new thinking) with seasoned professionals. They look to optimize the team to augment each other’s skills and balance out potential weaknesses.
We’ve all heard how AI is assisting with safety management. By analyzing historical safety data, AI is being used to identify when and where safety incidents are likely to occur before they happen. Cold weather is an indicator of the increased likelihood of slippages, material deliveries an indicator of material or equipment accidents, etc. AI helps project managers identify which subcontractors (or combination thereof) are most at risk so they can instruct foreman and superintendents to focus more attention on them. Combining historical data with images from cameras across the jobsite is also allowing another contractor to identify safety violations faster and organize more timely and targeted safety briefings to help reduce the risk of an incident.
Comparing data from a combination of project estimate, schedules, and a BIM model (among others), contractors are using helmet/body-mounted cameras and drone footage with AI to contrast actual project progress with planned progress. While the turnaround time for translating the camera data and comparing with planned production is time-consuming at present, the speed continues to improve. It allows project managers to know exactly where they need to spend time on the job site before they even arrive. This same information is starting to be used for identifying quality control issues as well.
Post Construction
Once the construction project is complete and turned over to owners, post-construction work falls to those contractors who provide service management. Here, AI is being used in a couple of different areas. Similar to the IoT data collected from construction equipment to predict when equipment is likely to fail, allowing proactive maintenance to be taken beforehand, data is also being collected from the equipment in buildings to predict when building equipment is likely to fail and provide proactive service management. Extending the capability even further, one small specialty contractor applied AI against service history data and was able to identify which parts were most likely to fail, ensuring that there were always available on service trucks, and which jobs always took longer, so that these could be scheduled in the afternoons. This approach allowed a larger volume of smaller turnaround jobs to be performed in the mornings. It significantly increased the amount of work a service technician could perform and made them much more efficient.
AI for the Back Office
Just like any other industry, one of the challenges for construction today relates to the labor workforce. Finding and retaining experienced employees is a key concern for many contractors. To assist with this, several contractors are experimenting with AI to predict the circumstances under which an employee is likely to quit. By using HR, ERP and PM data, they have found some of the early indicators an employee is thinking about leaving. These indicators include experiencing long periods of time between pay raises (or receiving low pay raises) compared to their colleagues, being on projects with frequent changes in staff and project scope and experiencing a lack of consistency in project processes across different projects. As they continue to analyze the data, additional indicators are being discovered.
What’s Next?
No discussion on the use of AI in today’s world would be complete without addressing ChatGPT. As with everything else, ChatGPT is making its way into the construction industry. It is being used for various tasks such as creating summaries of meeting minutes, including action items, due dates, and individual responsibilities. Additionally, it can answer questions about code requirements or potential safety hazards for projects, offering countless opportunities for the construction industry. AI capabilities are continuously being integrated into the software and applications used within the construction industry. Regardless of the size of the company or the available IT resources, contractors can start utilizing AI capabilities across multiple functional areas of their operations. AI is here today, and now we just need to leverage it.