Publications on ADS-B integrity, GNSS interference and UAS safety

Degradation of Automatic Dependent Surveillance-Broadcast integrity: Temporal clustering and anomaly persistence

Eugene Pik

Builds a global baseline of longitudinal ADS-B degradation from 229.0 billion state vectors and shows that kinematic anomalies are not independent random failures: they cluster in time for 98.9% of high-utilization aircraft, peak at a 3-second characteristic scale, and last a mean of 28.83 seconds, beyond the 5-second fault tolerance of decentralized fusion trackers. Recurrence across days is linked more strongly to network observability (hazard ratio 1.52) than to utilization (1.19). The paper argues for moving integrity monitoring from point-in-time anomaly detection to predictive system health management.

Author’s accepted manuscript, shared under CC BY-NC-ND 4.0 in line with Elsevier’s sharing policy. The version of record is at DOI: 10.1016/j.ress.2026.113534.

Predicting geopolitical instability through GNSS anomalies and air traffic data

Eugene Pik, João S. D. Garcia, Timothy A. Smith, Ibrahim Kocaman, Matthew Berra

Tests whether GNSS anomalies in ADS-B data can serve as early indicators of geopolitical events such as civil unrest and armed conflict, using OpenSky Network data and the ACLED conflict database. GNSS deviation density, normalized by traffic, is a positive and statistically significant predictor of conflict-event counts across all negative-binomial model specifications tested on 23,665 observations. The paper proposes open-source aviation data as an input to geopolitical risk assessment for aviation stakeholders.

Author’s accepted manuscript, shared under Springer Nature’s AM terms of use; not the version of record.

GPS Anomalies in Aviation: Preliminary Insights from Automatic Dependent Surveillance–Broadcast Data

Eugene Pik, Matthew Berra, Jeff Yearwood, Joao S. D. Garcia

An exploratory investigation of GPS gaps and deviations in a full year of global ADS-B data (376.9 billion messages), identifying 53,232 gaps in 37,546 flights and 5,878,275 deviations from 101,681 aircraft, and testing their association with 455,385 US and 30,160 international navigation NOTAMs. Anomalies concentrate where traffic is dense; gaps under active NOTAMs last on average 257 seconds longer; and many affected messages still carry a Navigation Integrity Category above 7. Solar flux showed no monthly correlation with gap counts.

© 2025 AIAA. Accepted manuscript, posted under AIAA’s self-archiving policy; the version of record is on Aerospace Research Central.

Utilizing artificial intelligence for National Transportation Safety Board unmanned aerial vehicle accident analysis and categorization

Eugene Pik, Joao S. D. Garcia

Uses a large language model to classify NTSB uncrewed-aircraft accident reports from 2006 to 2023 into standard occurrence categories and visualizes their geographic and temporal patterns. System and component failures not related to the powerplant, together with abnormal contact on landing, emerge as the leading categories, about 37% of the set. The approach is offered as a way to make unstructured accident narratives usable for safety policy.

Airport security: the impact of AI on safety, efficiency, and the passenger experience

Eugene Pik

A literature review of how artificial intelligence is changing airport security, from automated X-ray screening and biometrics to predictive threat analysis, and of the ethical, privacy and reliability questions those systems raise. Argues that transparency, accountability and continuous evaluation are conditions for AI to improve security without eroding trust.

Author’s accepted manuscript, shared under Springer Nature’s AM terms of use; not the version of record.

The Pilot Shortage: Implications, Repercussions, and Tried Solutions

Eugene Pik

An analysis of the global pilot shortage: its causes, its effects on airline operations, costs and safety, and the measures the industry has tried.

Published under CC BY-NC-ND 4.0 by the Hellenic Aviation Society.

Development of global regulations for Uncrewed Aircraft Systems – Europe and beyond: a continued survey and evaluation of progress

Christian Janke, Maarten Uijt de Haag, Eugene Pik

A survey of civil aviation authorities in 30 EASA member states, and beyond Europe, on the implementation of UAS regulation: remote-pilot certificates, registered aircraft, operator IDs, standard scenarios, PDRAs, SORA-based operations and Light UAS Operator Certificates. Compared with a 2022 survey, every measure grew, indicating expansion of the drone market and adoption of the EU regulation, with large differences between member states.

© 2025 IEEE. Accepted version, posted under IEEE’s author posting policy. The published version is on IEEE Xplore.

Detecting GPS Anomalies in Aviation Using ADS-B: Correlating Coordinate Gaps and GPS Deviations with NOTAM Warnings

Eugene Pik, Matthew Berra, Jeff Yearwood, Joao S. D. Garcia

The conference version of the GPS-anomaly study later published in the Journal of Aerospace Information Systems: methods for detecting gaps and deviations in 376.9 billion ADS-B messages and correlating them with navigation NOTAMs, presented as first author.

© 2024 the authors. Published by AIAA with permission.

Empirical Propeller Mass Sizing for Small-Scale Aircraft

Amy Chang, Eugene Pik, Robert W. Deters, Or D. Dantsker

An empirical model for estimating propeller mass in small uncrewed aircraft, where propeller weight has an outsized effect on weight and balance, endurance and manoeuvrability.

© 2024 the authors. Published by AIAA with permission.

Global Traffic-Normalized Baseline for ADS-B Kinematic Integrity and Ghost Records

Eugene Pik, João S. D. Garcia, Timothy A. Smith

Establishes a global, traffic-normalized baseline of ADS-B kinematic consistency from 246.7 billion messages and 77.5 million flight hours in 2025 across 260 Flight Information Regions. Introduces the Global Envelope Thresholding (GET) framework, which removes the survivorship bias of legacy filtering pipelines and identifies 3,818,743 "ghost records", a newly defined class of severe anomalies with partially missing telemetry. Negative-binomial regression reveals substantial regional disparities in surveillance consistency. The benchmark dataset and a Python API are open.

The SSRN preprint (April 2026) appeared under the title “The Invisible Infrastructure: A Global Baseline for ADS-B Kinematic Integrity and the Ghost Record Phenomenon”; the PDF here is the revised manuscript.

Cross-Channel Anomaly Coupling in ADS-B: A Surveillance-Integrity Monitor for Air Traffic Management

Eugene Pik

Analyses one year of OpenSky state-vector data (117.5 million aircraft-hours over 240,001 airframes) and shows that severe longitudinal-acceleration, vertical-acceleration and track-angle-rate anomalies are not independent: longitudinal anomalies are 3.14 times more likely when vertical anomalies are present, and the coupling violates conditional independence. The paper turns that structure into a deployable surveillance-integrity monitor for air navigation service providers.

• GET-25: The Global Spatiotemporal ADS-B Kinematic Consistency Benchmark (2025 data, Python API)
• Degradation of ADS-B integrity: temporal clustering and anomaly persistence (replication data and code)
• Multi-dimensional coupling of kinematic anomalies in ADS-B surveillance telemetry (replication data and code for the cross-channel coupling study)
• Hourly Global ADS-B Kinematic Sensitivity Dataset (2025)
• Geospatial Dataset of GNSS Anomalies and Political Violence Events (2023)
• 2023 GPS Anomalies, NOTAMs, and Aircraft Traffic, with a hexbin aggregate
• SQL and Python scripts for OpenSky Trino analysis of GPS anomalies
• GPT-4 assisted categorization of NTSB UAV accident reports
• Commercial and Recreational Active UAV Fleet in the U.S. 2016–2022

All datasets and code →