The datasets were designed to monitor and predict stress levels among operators in an industrial environment through the use of smartwatches. The structure is hierarchical and scalable, allowing for a transition from a small operating unit to an entire corporate infrastructure. Reference to the application: https://data.d4science.org/ctlg/ResourceCatalogue/incop_-_industry_5_0_collaborative_platform

1. Basic Configuration: 8 Operators and 3 Departments This is the standard model (contained in the file dati_8operatori_3reparti.json). The dataset is organized into time blocks, i.e., measurements taken every 5 minutes for 3 working days, and each record contains:

Shift Assignment: Assign each device (deviceId) to a specific department and work area. Oven Department Production Department Packaging Department Telemetry: Reports the stress level detected in real time for each operator. Predicted Stress: Includes stress values calculated by a predictive model (stress_pred) to anticipate potential critical situations.

1. Model Scalability To test the robustness of the system and its capacity for large-scale analysis, the same logical structure of the dataset was replicated and expanded in three incremental scenarios:

Medium Configuration (80 Operators - 6 Departments): Extends monitoring to an entire plant, doubling the number of departments to include logistics and maintenance areas.

Large Configuration (800 Operators - 9 Departments): Simulates a complex production hub, where data flow requires advanced management of simultaneous telemetry.

Enterprise Configuration (8000 Operators - 12 Departments): Represents the maximum load scenario, useful for testing artificial intelligence algorithms on Big Data and for monitoring multiple interconnected industrial sites.