Aplicação de inteligência artificial na gestão de lições aprendidas em projetos de engenharia, suprimentos e construção no setor elétrico

Abstract

Engineering, Procurement, and Construction (EPC) projects, particularly in the electricity sector, involve high technical complexity, multiple stakeholders, and a large volume of information generated throughout the project lifecycle. This research addresses the challenge of the systematic loss of technical knowledge in Engineering, Procurement, and Construction (EPC) projects in the Brazilian electricity sector, resulting from informality in capturing lessons learned, information dispersion, and the absence of standardized processes. Given this problem, the objective of this study was to develop and validate a technological conceptual framework based on Artificial Intelligence (AI) capable of automating the capture, categorization, and reuse of lessons learned, promoting the systematization of organizational learning and the continuous improvement of project management. The research adopts a qualitative approach of an applied nature in organizations within the electricity sector. Data collection involved document analysis and semi-structured interviews with project managers, engineers, and AI professionals. The empirical analysis was conducted using the Gioia Method, allowing the construction of inductive categories organized into a four-phase structure in light of theoretical frameworks on knowledge management, organizational learning, lessons learned frameworks, and AI applications in project environments. This allowed for the identification of critical barriers and the translation of these findings into functional and technical requirements for the artifact. The resulting framework integrates technological components (NLP and Machine Learning), standardized operational flows, and organizational implementation guidelines. As a contribution, the study proposes a functional artifact, validated by experts, that can be replicated in different organizational contexts within the EPC sector. Theoretically, the research fills a gap in the literature by articulating knowledge management, organizational learning, and emerging AI technologies. From a practical standpoint, it offers an innovative solution to reduce rework, preserve critical knowledge, and enhance maturity in project management in the Brazilian electricity sector.