Hey everyone, good afternoon. I work with logistics warehouse construction in Brazil and one thing that always bothered me is the enormous effort required to structure a WBS. With that in mind, I worked together with Claude to build a skill that solves this problem. I basically taught it all the construction logic we use day to day — longitudinal advancement, quadrant division based on bay counts, floor slab sequencing with dock border strips poured last, items that are independent of docks like fire doors and louvers, and even L-shaped warehouse geometries. Everything aligned with AACE International Recommended Practices. The skill starts from a questionnaire of about 30 questions about the project and generates the complete WBS, standardized and ready to import into MS Project or Primavera. The practical result is that a task that used to take me around 48 hours as a planner now takes about 30 minutes between answering the form and validating the output. I've already tested it on warehouses from 12,000 to 44,000 m², crossdocking and single-dock layouts, with different configurations of annexed buildings and external areas. The most interesting part is that Claude doesn't replace the planner's expertise — it learns and replicates the logic we normally carry only in our heads. The next step is building a second skill for activity logic (predecessors and successors), turning the WBS into a fully sequenced schedule. If anyone is interested in discussing this further, I'm happy to connect.

Please how can I download Excel material on EVM. I have watched the Youtube tutorial, very interesting

Prompt used in construction

I'm outlining a video on what I see as realistic predictions about AI in construction. I'm keen to hear people's thoughts/criticisms So first, a qualification. Technology occurs in step changes. ChatGPT was released and overnight everything changed. But then air travel hasn't changed much in 50 years. So I think the only valuable predictions we can make are about how current tech will change things. I think predictions become much less useful when we make assumptions about new tech. PREDICTION 1 - AI is not going to change anything fast Outsourcing has been around for 20 years. Construction companies can get amazing quantity surveyors, admins, etc. in countries like the phillipines for a fraction of the cost. And it hasn't changed the job market in Aus/US/UK, etc. Outsourcing is basically like having AGI. PREDICTION 2: The Hallucination Problem Is Worse in Construction Than Any Other Industry Construction is not fundamentally about being "super smart" its about not making mistakes. A 5% error on a $10m job is a $500k mistake. Saving a bit of time putting the estimate together is not worth the risk. If einstein produced your construction estimate. I dont think it would make it more accurate. Its more about getting quotes, doing due diligine, double and triple checking. Construction isn't an intelligence problem PREDICTION 3: Robotics Are a Very Long Way Off The tehcnology to do driverless cars has more or less been around for 15 years. But we still don't have them. That's because of regulation, edge cases etc. Construction is full of these issues. Plus industrial relations and unions will make it harder. PREDICTION 4: Software Will Matter More Than AI Itself I dont think construction companies will be building custom AI automations, etc. Or maybe they will. But they will largely rely on vendors who have them built in. Therefore, using software and capturing the benefits of good AI use cases will be more important then actually using AI. PREDICTION 5: General Contractors Will Get Squeezed