Modern industrial systems generate large volumes of data from sensors and automated equipment, which are widely used in machine learning research. However, managing these datasets — particularly in terms of organization, documentation, and sharing — remains a significant challenge for research teams. The lack of consistent data management practices can hinder the reproducibility of experiments when data sources or applied transformations are not clearly documented.
This thesis presents the design and implementation of secure architecture aimed at centralizing and governing industrial datasets using structured metadata. The data considered primarily originates from real-time supervision and acquisition systems deployed in the renewable energy sector. The proposed solution relies on a relational database, secure application programming interfaces for access control, and the integration of a JupyterHub environment dedicated to data preparation and exploratory analysis.
The objective of this architecture is to improve the organization, sharing, and reproducibility of industrial data used in research, while minimizing operational maintenance requirements.
| Date | 23 Apr 2026 |
|---|
| Original language | French |
|---|
| Awarding Institution | - École de technologie supérieure
|
|---|
| Supervisor | Antoine Tahan (Supervisor) & Pavel Guerra Côté (Co-supervisor) |
|---|
Nasr, C. (Author),
Tahan (Supervisor) & Guerra Côté (Co-supervisor),
23 Apr 2026Student thesis: Master's thesis › Master in Engineering: Information Technology Engineering