3 ms·

Anomaly detection of private jet flights

• kcimc 5mo ago
  i built a small system that tracks a fixed cohort of business jets (from FAA registry data, matched via ICAO hex) and counts how many are airborne at a given time. it ingests ADS-B exchange heatmap files (30 min intervals), backfills historical counts, and compares the current value to a baseline for similar times of day/week. the output is a simple deviation, shown as an "emergency level". in practice it mostly shows a strong daily cycle, with occasional spikes during holidays, or on april 6th when trump threatened "a whole civilization will die tonight".
  • saargrin 5mo ago
    where did they go on april 6th btw? would be nice if you had like a historical record of what happened when your alert level went above 3
• pcdoodle 5mo ago
  [dead]