Problem Specification Summary

This project analyzes historical data from the Barcelona bike sharing system to identify the most critical timeslot (weekday--hour) for each station.
A station is considered critical when its number of free slots is 0, meaning it is completely full.

The analysis uses two datasets:

• register.csv: timestamped readings of used and free slots for each station
• stations.csv: station coordinates and names

The Spark application: - Computes criticality for each (station, timeslot) - Filters timeslots above a minimum criticality threshold - Selects the most critical timeslot per station (with deterministic tie‑breaking) - Outputs results into a single .kml file for map visualization

---

Cluster Configuration Used

The Spark application was executed on the Polito BigData Cluster using:

• Server Options: Bigdata Education Reserved
• Allocated Resources:
  • 4 CPU Threads
  • 16 GB Memory
• Maximum Possible Resources:
  • 16 CPU Threads
  • 32 GB Memory

---

Final Output

After executing the Spark job, the system generated a KML file (see Out_data/output.kml). for visualizing the most critical stations you can upload the kml file to https://kmlviewer.nsspot.net and see the map.
The final output image is available in the main directory as:

This map highlights stations that frequently reach full capacity, along with their critical timeslots.