
Is your data ready for AI?
Completeness, consistency, standardization and traceability: the four questions that come before any model.
The talk I presented at the SMRP 2026 conference: how generative AI enters maintenance planning, decision-making and failure analysis, and what needs to be in place before it.

The talk follows the P-200, a pump with five sensors and a history that only says “done”. Each chapter answers a question every maintenance manager has asked.

Completeness, consistency, standardization and traceability: the four questions that come before any model.

One detects and predicts; the other interprets, connects and explains. Complementary, not competing.

How to connect AI to the maintenance system with permissions, context filtering and an audit trail.

Express service requests, assisted maintenance, reports, weekly scheduling, failure patterns, diagnosis and governance.

Thirty years in maintenance and asset management, from the shop floor to founding a maintenance management platform with more than 250 customers in Brazil and the US. At SMRP 2026, I shared the stage with Rafael Araujo Kluska, PhD, a generative AI specialist: two sides of the same problem, the reality of maintenance and the capability of AI.
The talk is the beginning. At rodrigorotondo.com.br I bring together the method for AI to work like an experienced maintenance and reliability engineer, with evidence and with the decision in the hands of whoever signs off.