Predictive maintenance
SMPIA predictive-maintenance architecture
A graduation project exploring a microservices-based maintenance platform with a foundational machine-learning integration.
- Client
- Instituto Profesional INACAP
- Period
- 2024
- Engagement
- Academic project
- Visibility
- Academic project
Context
Project overview
The project examined how operational maintenance data, application services, and a prediction component could coexist behind a coherent product interface.
Challenge
Challenge
It required adopting a multi-service architecture and connecting a basic TensorFlow/Keras workflow without losing clarity in the broader application design.
Responsibility
Responsibility
- 01
Design the service boundaries and shared data flow.
- 02
Build the React interface, Strapi service, FastAPI component, and PostgreSQL layer.
- 03
Integrate a foundational predictive model into the application workflow.
Approach
Decisions and approach
- 01
Separated application content, prediction concerns, and the user-facing workflow into explicit services.
- 02
Used the project to evaluate integration boundaries rather than presenting machine learning as a standalone feature.
- 03
Kept the predictive component intentionally foundational and proportional to the academic scope.
Outcome
Outcomes
- Delivered a complete microservices architecture for the graduation project.
- Demonstrated a working path between the application and a basic ML component.
- Established a system design that could support deeper experimentation later.
Technology