Los Angeles-based programmer and media artist Kyle McDonald has launched the Apocalypse Early Warning System website. The online tool tracks private jet activity in real time to predict global threats based on the behavior of the world’s wealthiest people.
Los Angeles-based programmer and media artist Kyle McDonald has launched the Apocalypse Early Warning System website. The online tool tracks private jet activity in real time to predict global threats based on the behavior of the world’s wealthiest people.
According to the author, since the richest people in the world have direct contact with government structures and are the first to receive inside information about threats to civilization, the anomalous increase in the number of private flights may indicate their attempt to urgently evacuate to their private estates and bunkers before the onset of a global catastrophe.
The developer emphasizes that the AEWS system is not designed to detect missiles or physical threats that are already in the air and approaching targets. Instead, the tool focuses on human factor analysis. The greatest value is in proactive actions — the anomalous behavior of individuals or organizations that react to classified information hours or even days before it becomes common knowledge to the general public.
The AEWS project is built on the collection and processing of open source intelligence (OSINT) in several stages:
Data sources: The tracker forms a global metadata database, combining aircraft records from the ADS-B Exchange aviation resource, Mictronics/tar1090 platforms, and the official FAA registry by unique ICAO identifiers.
Getting coordinates: Flight data is imported from ADS-B Exchange heatmap files, which are updated every half hour. The backend loads the most recent heatmap, compares it to our database of business jets, and captures the coordinates, altitude, speed, heading, and flight status of each aircraft. Military aircraft are only displayed if the user activates the corresponding toggle.
Anomaly Calculation (Sigma Index): The system analyzes historical data stored in SQLite. The panel compares the current number of aircraft in the air with the baseline for the same day of the week and time.
The model also takes into account US public holidays so that pre-holiday activity is not perceived as an alarming spike. The difference between the actual and expected number of aircraft is converted into a sigma index, which minimizes the impact of random night-time fluctuations. At the output, the system assigns a threat level from 1 to 5.
The service also calculates the approximate total seating capacity in the air by comparing business jet models with their maximum passenger capacity.
McDonald notes that peak activity is not a 100% guarantee of disaster, but the system has already proven its worth. The highest activity index was recorded on April 6, during the escalation of the conflict in the Middle East, when Iran attacked American and Israeli facilities.
«It scared me. I remember thinking, 'Oh my God, this is really happening,'» the developer shared.
At the same time, McDonald emphasizes that even the maximum level of five only means that the current indicator is an extreme positive deviation for the model.
Such a spike could still be caused by holidays, major sporting or political events, data artifacts, or technical collection errors. The site has specifically maintained an archive so that users can see these historical false positives and assess the system’s error.
