Digital Exclusive: Ten ways AI is expediting construction projects in the processing industry
B. GALLAGHER, Graycor, Oakbrook Terrace, Illinois (U.S.)
Like most construction megaprojects, processing projects have scaled up in complexity while offsetting that complexity using tools such as computer-aided design (CAD), building information modeling (BIM) and project management software. The bad news for the processing industries is that complexity has been outpacing the implementation of productivity tools—however, this may be about to change.
As early as 2014, the Project Management Institute (PMI) recognized that “…construction and engineering projects have become more complex and ambitious faster than our ability to manage them. Oil/gas/infrastructure projects now are much longer in duration and far more complex than even ten years ago, with concomitant increased risks and failures.” In the intervening decade, the intricacies of energy and processing megaprojects have only increased (FIG. 1). Adding to the complexity of these types of projects are a high regulatory burden and a particular focus on environmental impacts.
FIG. 1. Projects with a long construction timeline, such as the Sherwood VIII cryogenic gas processing plant (shown here), benefit from productivity tools that enable communication and coordination between project teams, subcontractors and suppliers.
Changes to technology tools commonly used in construction, however, were happening only incrementally until artificial intelligence (AI) caught the attention of the public—and of various industries—in 2022. Since then, design and construction teams have begun embracing its potential for capital projects.
The exponential improvement offered by AI rests on its ability to use advanced algorithms and real-time monitoring to dramatically address and mitigate multiple risk categories. Complementary to AI is machine-learning (ML): whereas AI can perform some tasks that previously demanded human effort, ML enables software to predict outcomes with good accuracy. Software companies have made great strides using artificial neural networks with interconnected nodes that mimic biological neurons and their ability to adapt and learn. ML uses this type of pattern recognition to create models that can then develop iteratively as new data is fed into the system, thus operating without the need for explicit programming.
In addition to understanding how AI tools can enhance construction efficiency, it is important to understand when to begin incorporating them. Further findings of the PMI report showed that decisions and management practices commencing after the construction phase had much less effect on project outcome than those made during preconstruction. Bringing all stakeholders, including engineers and contractors, to the table as early as possible helps eliminate uncertainties that might cause a large ripple effect and negatively impact project outcomes. Therefore, many of the 10 key areas in which AI can improve project delivery begin during project planning.
#1: Feasibility studies and site selection. Predictive analysis of large data sets using AI can reveal findings on multiple factors that influence site selection. This includes geological data and terrain features, both of which are determinants of the site work and structural designs that will be required. AI tools can assess existing environmental conditions, benchmark them against regulatory requirements, and even identify existing conditions that could affect resilience in the face of natural hazards.
Regional workforce skills, demographic trends and local economic conditions are other factors that should be considered during site selection. While experienced site selection teams have long examined these variables, by using AI they can consolidate increased types and amounts of available information. For example, in addition to providing insights into existing regional workforce skills, AI tools may be able to thoroughly assess local training options.
#2: Design-phase decisions. Generative AI, especially the iterative, pattern-recognizing algorithms of ML, can provide more design options and alternatives than would otherwise be feasible. With the complex layout and fittings within process facilities, this is of great benefit. AI tools can also minimize clashes within the facility’s complex layouts, working in conjunction with BIM models. Clash-detection AI tools can analyze operational characteristics as well as physical ones, identifying conflicts between systems. These tools can also respond to constraints input by designers and other users.
#3: Cost estimation. Financial and market data can be fed into AI tools to provide a reliable “big picture.” The tools can also estimate construction costs using data sets on material and labor costs, transportation costs, regulatory requirements and more. As with other AI-enabled tools, cost calculators can go beyond typical estimation programs to capture a substantial number of additional variables, such as differences based on contract types or historical data that is refined to account for seasonal or other time-dependent fluctuations. Future operating costs can also be estimated using AI, since many AI models can predict future energy prices and consumption patterns based on historical data and market trends.
#4: Scheduling. Like its application in cost estimating, AI can streamline scheduling by forecasting not only potential delays due to labor and materials availability, but also other market fluctuations, based both on historical scenarios and iterative processes that incorporate new data. AI tools can perform sophisticated analyses of dependencies, resulting in greater accuracy and enabling teams to begin risk-mitigation efforts sooner than ever before. AI tools can even be used to actively troubleshoot material sourcing bottlenecks.
Jobsite productivity can be enhanced with AI by conducting scans of site and subcontractor progress. AI tools can analyze where on the site workers could be moved to expedite progress and optimize available labor. Real-time schedule progress can be tracked against projected schedules using AI, again allowing risk remediation strategies to occur in a timely manner. AI can even support this risk mitigation by running “what-if” scenarios of various scheduling changes.
#5: Automation and robotics. The use of automated tools spans all project phases, from planning and preconstruction through construction and operation. AI-enhanced drones and intelligent robots can perform detailed site surveys. Once construction has begun, they can identify safety hazards and monitor equipment performance, as well as conduct site inspections using sophisticated methods that may catch things human eyes cannot as easily detect. This is especially helpful for energy and process building sites, where safety can be a particular concern. In some cases, robotics-enhanced offsite assembly may result in greater efficiencies. All AI-enhanced assembly tasks, whether on or offsite, are particularly valuable in the current environment of limited labor availability.
#6: Sustainability. Achieving greater sustainability in construction relies on implementing good metrics and effective tracking systems. AI can help teams identify the most sustainable materials and methods, then help them track the carbon content of materials used. This can help reduce waste, allocate resources and minimize transportation-related inefficiencies, all of which reduce overall emissions. AI tools can even use predictive analytics to assess a construction project’s potential impact on the environment.
#7: Project management and constructability. AI-driven collaboration platforms, in conjunction with estimating and project management tools, enable communication and coordination between project teams, subcontractors and suppliers, thereby increasing transparency and productivity. Natural language processing (NLP), a subcategory of AI in which computer programs can interpret human language, allows for the quick extraction and processing of data from construction documents. Construction management software often incorporates AI-based processes that automatically compare, for example, actual progress to planned progress, or propose project delivery alternatives.
#8: Quality. AI tools can analyze data to identify patterns related to quality issues, allowing for timely interventions to maintain high standards throughout the construction process. The use of AI-enhanced tools to augment human inspection schedules provides a particular boost to quality, improving error detection and managing cost and labor considerations.
#9: Safety. In addition to AI-enhanced robotic tools that improve the safety of inspection routines, various cameras and sensors are available that can scan for safety violations as they occur on the site. Analytic tools can be used to determine which site activities are the most likely to result in an accident, paving the way for active risk mitigation.
#10: Performance monitoring and predictive maintenance. Process facilities comprise a heavy concentration of sophisticated and critical components. AI-enabled sensors can monitor—in real time—the health of these components. Algorithms analyze sensor inputs and alert facility personnel to potential problems before they occur, allowing for the replacement of components before they fail.
Takeaways. The time to plan for AI-enhanced plant maintenance is during the facility’s planning and design stages, so all appropriate systems and technologies can be designed. Projects with a very long construction timeline, such as those in the process industries, require significant investment in up-front planning, design and development. AI tools, in the hands of an experienced team of contractors, engineers and other design partners, are now helping offset those heavy investments.
ABOUT THE AUTHOR
Brian Gallagher is the VP of Corporate Development for Graycor and has more than 30 yr of experience leading strategic planning, organizational development, marketing and sales efforts for design and construction firms. He is author of three books and was named a Top 20 Construction Influencer by Procore. The author can be reached at Brian_Gallagher@graycor.com.
Comments