Artificial Intelligence in Construction
Many companies are developing AI tools to improve the building process for contractors, from estimating and managing finances to monitoring sites and reviewing contracts. But even when a tool is made specifically for construction, it is still important to know how to get what you want from it. Using AI isn’t about learning how to code; it’s about clear communication. The following recommendations are intended to help construction professionals ask and receive the best answers from AI and are especially useful when using AI tools for estimating.
5 Tips for AI Construction Estimating
1. Define Inclusions & Exclusions
Clearly state what trades or scopes are part of the bid and what aren’t.
“Include drywall and paint; exclude flooring.”
2. Upload Documents
Always provide the plans, schedules, and/or scope of work. The more context the AI has, the better the output.
3. Outline Assumptions
If plans are vague, tell the AI what to assume.
“Assume standard 4-inch slab on grade.”
4. Be Specific with Edits
When asking for changes, use precise language; so instead of “fix the price,” say:
"Increase unit cost for framing by 10%.”
5. Output Format
Clarify what you want your output to look like.
“For all scopes, please provide separate items for labor and materials.”
AI Best Practices
It is also helpful to keep in mind a few best practices for interacting with any AI beyond the construction-specific advice above.
- Give Guardrails: You must provide guardrails for what it should or shouldn’t do; without clear guardrails, the AI will not know what you are asking for. Think about AI as if it were a person, they won’t understand exactly what you want if you send them a vague one-line email; neither will the AI.
- Be Clear and Accurate: If the language you are using is not clear, has spelling mistakes, or contradicts itself, that will make it hard for the AI to know what you are looking for. Re-read your text to be sure it is clear what you are asking for.
- Provide a Role: Giving the AI a specific perspective helps it deliver more relevant and accurate results. When the AI knows which role to assume, it can tailor its response based on the responsibilities and viewpoint of that particular job or person. The more specific you are about the role, the more useful the response will be.
