With the potential to redefine how people around the globe live and work, Artificial Intelligence (AI) is the most significant technological innovation of our time. Not only do these tools promise to unleash a new wave of human productivity, but they have the power to obliterate human limitations and expand our existing conceptions of the possible. The AI revolution, still in its infancy, is already impacting every sector of the economy, including the construction industry. From planning and estimating through procurement and delivery, AI has applications for a wide variety of construction activity.
However, construction professionals must be cognizant of the fact that not all AI tools are the same, and careful about the ones they select to help them work more effectively. In this post, we discuss several uses of AI for construction and contrast two AI models – generic and contextual – to help you make informed decisions about what to look for in AI-enabled solutions.
What Are AI Applications for Construction?
AI has the potential to make every step of the construction process smarter, faster and more efficient. Here are several uses of AI for construction.
- Site-Walk Data Collection: Site walks are a crucial part of project planning and execution. Yet with so many stakeholders sharing so much information so quickly, it’s easy for details to fall through the cracks. The use of talk-to-text AI tools can capture notes, specifications, useful nuggets of information and other details that are currently easy to miss, resulting in scopes of work and cost estimates that are more accurate, and less back-and-forth between all parties. Site-Walks are also great opportunities to capture images of project-related assets and specifications. The use of AI to auto-generate descriptions of pictures and images can shave hours off of data management and enrichment tasks.
- Streamlined Scope of Work Creation: The same technologies that can capture site-walk data can synchronize with back-office software for faster and more informed Scope of Work generation. Not only do these tools eliminate manual data entry, they ensure each SOW is built in a consistent format, for easy breakdown by trade, material or contractor specialty. AI-enabled SOW generators can quickly consume notes and project details (including linked assets) and then “chat” with the user to answer questions and validate assumptions in order to rapidly generate a comprehensive Detailed Scope of Work.
- ROM/Conceptual Estimating: Currently, an estimator may spend days developing early-stage, Rough Order of Magnitude (ROM)/conceptual estimates. AI can reduce that timeline to mere minutes, in part by automatically interpreting troves of project documents, including specs, RFPs and Scopes of Work. Users can interact with chat agents to ensure AI understands the context of the project it is estimating. AI reads all the project-related documentation and interprets the scope of the project before conversing with the user to fine tune the scope of the requested estimate.
- Line-Item Estimate Building: AI agents can assist in line-item estimate creation by suggesting alternative materials that maximize budget without sacrificing quality, provide notes explaining the rationale for costs and quickly restore previous versions of an estimate when necessary. When determining which line-items to include in the estimate, AI can also provide a basis for including any line-item, essentially providing a link to the specific project task that triggered the need to include that line-item in the estimate. This significantly streamlines the estimate creation and review process.
This list is by no means exhaustive. There are currently even more AI applications than those included here, and there will be even more in the future. Artificial Intelligence is here now, and every indication is that it is here to stay.
Introducing Flash AI Estimating, an AI model built specifically to create reliable early-stage estimates in just minutes. Learn more here, and start estimating in a Flash!
Generic AI vs. Contextual AI
Given that we’re in the early stages of adopting the technology, it’s tempting to think of AI tools as monoliths – that AI is AI is AI. But just like different models of drills and saws, different AI models have distinct uses and features, and the contrasts between them are significant. Next, we look at generic and contextual AI models, and explore which is a stronger choice for use in construction.
Generic AI
ChatGPT. Copilot. Claude. People around the world turn to these innovations every day for quick answers to questions, meeting summaries, outfit suggestions – you name it. Generic AI models are trained on massive, general-purpose datasets, making them great choices for low-risk tasks. While personal use of general-purpose AI tools is still skyrocketing, there’s been recent “softening” when it comes to corporate use of generic/public AI where security of intellectual property is a concern. Many companies are limiting use of public tools that exist outside the corporate firewall and are instead implementing their own instances of general-use AI that’s limited to accessing only their internal data and documentation.
Contextual AI
In contrast to generic AI, contextual AI models are purpose built for specific uses. They are trained or configured using the language, logic and data of the user’s business. Contextual AI models integrate into and enhance existing workflows, making processes more efficient. Furthermore, these powerful, custom-built tools can learn the language, rules and sensibilities of niche markets to deliver practical, outcome-driven intelligence. The beauty of utilizing contextual AI models is that you can essentially create teams of AI agents trained in specific tasks or parts of a workflow. Any given agent can complete its specialized work and then pass the result to another agent with a different skillset. Also, when using a contextual AI agent, any information shared with AI stays within the contextual confines of that specific conversation.
Which AI Model Is Appropriate for Construction?
Given its enhanced powers and application for specific purposes, contextual AI models are absolutely appropriate for construction work. These tools can augment human expertise by replacing manual, time-consuming tasks, creating more consistent and efficient workflows. Plus, contextual models can make suggestions and decisions based on in-house data and documentation, making them more reliable than AI models that pull information from unknown sources and tend to “hallucinate” incorrect data.
Artificial intelligence is going to revolutionize how construction gets done. It already is. Yet, you’d be wise to think judiciously before adopting an AI tool. Contextual AI models can help reduce risk, increase accuracy and strengthen how projects and facilities are planned, managed and executed. Generic AI models cannot. Choose wisely.
An AI Model Built Specifically for Estimating
The natural follow-up question to everything we’ve discussed so far is, “So who is making contextual AI models for construction?” Well, Gordian has recently launched the world’s first AI-powered construction estimating assistant: Flash AI Estimating. What separates our model from others is that it has been rigorously trained for the task of creating quality estimates and draws from our proprietary RSMeans Data, the gold standard cost estimating database in the industry. Flash AI Estimating reviews your construction documents and chats with you to rapidly generate early-stage cost estimates powered by RSMeans Data. Upload your docs, chat with Flash AI Estimating and quickly produce traceable, defensible estimates for construction work in a matter of minutes, rather than hours or days.
Interested in seeing Flash AI Estimating in action? Request a demo today!
