EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Desenvolvimento de sistema prescritivo de manutenção industrial: análise de viabilidade técnica e eficácia de algoritmos python na antecipação de falhas

Elton Ribeiro Bastos, Ronaldo Gomes Figueira

Revista ft · 27.05.2026 · 30 (158) · 01-48

DOI: 10.69849/dv480h14 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Revista ft, online erschienen am 27.05.2026. Verfasst von Elton Ribeiro Bastos und 1 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Achieving the stage of prescriptive maintenance is the goal of areas seeking maximum autonomy, where the system not only predicts the defect but dictates the solution even before the stoppage occurs. However, a real obstacle in factories is the lack of accurate records and breakdown histories, which ultimately hinders the training of prescriptive models. This research shows that this gap can be overcome by using synthetic datasets, such as the Microsoft Azure dataset, to prepare algorithms safely and effectively. The work uses the Python ecosystem as a choice for data analysis, leveraging its dedicated tools for this purpose. The central contribution is a proposed hardware integration that includes Siemens drivers (S7-1200/1500) and Advantech gateways (ECU-1051). This structure was designed for real-world field variations that express the pattern of the data used in the tests, ensuring that the Ensemble LSCP algorithm accurately identifies signs of wear. This choice allows logic tested in a virtual environment to be applied in the future in a real-world environment similar to the one proposed. The results prove that this combination of industrial equipment and open-source software is a financially viable way to avoid unwanted interruptions in the production flow. By automating prescriptions, the system removes the burden of human diagnosis and accelerates decision-making in Maintenance Planning and Control. It is concluded that the model is practical and scalable, offering a safe roadmap for companies that wish to upgrade their operational availability.

Zitieren

APA

Bastos, E. R., & Figueira, R. G. (2026). Desenvolvimento de sistema prescritivo de manutenção industrial: análise de viabilidade técnica e eficácia de algoritmos python na antecipação de falhas. Revista ft, 30(158), 01-48. https://doi.org/10.69849/dv480h14

Vancouver

Bastos ER, Figueira RG. Desenvolvimento de sistema prescritivo de manutenção industrial: análise de viabilidade técnica e eficácia de algoritmos python na antecipação de falhas. Revista ft. 2026;30(158):01-48. doi:10.69849/dv480h14

BibTeX
@article{Bastos2026,
  author = {Bastos, Elton Ribeiro and Figueira, Ronaldo Gomes},
  title = {Desenvolvimento de sistema prescritivo de manutenção industrial: análise de viabilidade técnica e eficácia de algoritmos python na antecipação de falhas},
  journal = {Revista ft},
  year = {2026},
  volume = {30},
  number = {158},
  pages = {01--48},
  doi = {10.69849/dv480h14}
}
Kostenlos starten →