East Texas A&M Tests Pretrained AI Models to Detect GPS Spoofing in Drone Flights

Published
Source
Dronelife.com
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 RobotToday Industry Briefing RobotToday Industry Briefing graphic: a blueprint-style humanoid robot outline with annotations for perception, actuation and mobility, next to the Industry Briefing wordmark. PERCEPTION ACTUATION MOBILITY ROBOTTODAY Industry Briefing Editor-curated robotics news from around the world HUMANOIDS/ INDUSTRIAL / DRONES / AI & RESEARCH RobotToday robottoday.com/industry-briefing
RobotToday Industry Briefing

Researchers at East Texas A&M University (ETAMU) tested whether AI models can detect GPS spoofing by identifying inconsistencies in drone motion data, without using GPS readings. The study analyzed a public dataset of 96 real flights: 42 normal flights and 54 conducted under controlled spoofing attacks. Researchers withheld 22 complete flights from training to assess performance on unfamiliar data. The work was accepted for presentation at WATCH 2026, a peer-reviewed workshop held alongside the Association for Computing Machinery’s Conference on Computer and Communications Security in The Hague, Netherlands. National Science Foundation Award No. 2502117 funded the work.

The approach uses onboard sensor data about speed, direction, and elapsed time, in a manner Jinoh Kim, an ETAMU computer science professor, compared with dead reckoning. The team compared eight models, including models trained from scratch and pretrained models. Models pretrained to recognize patterns in numerical time-series data performed best, including when data was limited. With 10 consecutive readings, the Chronos model scored 97.5 percent on a measure balancing detection of abnormal sequences against alert correctness. The findings indicate that motion data may help flag unreliable location information, which could matter when spoofing risks misdirecting military or civilian drones.

The evaluation covered recorded flights only; according to the university, testing the checks during flight is the next step. The broader project, supported by a two-year NSF grant of about $200,000, also explores how a drone could estimate its location when reliable data is unavailable. The source does not disclose a timeline for in-flight testing.

Original report

East Texas A&M Researchers Test AI to Detect GPS Spoofing in Drones

Dronelife.com · Ian McNabb
Read at Dronelife.com

More related news

  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8
  9. 9
  10. 10
  11. 11
  12. 12

Related Suppliers

Roboris Srl

Pisa, Italy software developer of Eureka Virtual Machining for offline robot programming and CNC simulation; offices in five countries.

CSP GmbH & Co. KG

German industrial software maker (est. 1991); CSP Manufacturing OS for quality assurance, process data, and screw-joint testing in discrete manufacturing.

Industrial Manufacturing Software & Algorithm Provider System Integrator

Regulus Cyber

Israeli GNSS cybersecurity firm (2016, Haifa); Pyramid software blocks GPS spoofing on autonomous vehicles; Ring R1 C-UAS system defeats rogue drones.

Commercial Service Robots Autonomous Vehicle / UGV Human-Machine Interaction

Indra Sistemas, S.A.

Spanish defence and technology group (Indra Group) behind the ARMS, Crow and ARACNE counter-drone systems and the TARSIS family of military drones.

SkyfireAI

US-based drone AI company delivering turnkey Drone First Responder (DFR) and multi-ship autonomous UAS solutions for public safety and defense.

Perceptual Robotics

Bristol university spin-out building fully autonomous inspection drones and AI software for wind-turbine blade inspection.

Skyline Nav AI

Cambridge, Massachusetts firm founded 2020 making Pathfinder camera-and-IMU navigation for GPS-denied drones and vehicles; holds an AFWERX SBIR Phase II award.

Kutting Technology (Suzhou) Co., Ltd. 苏州酷丁科技有限公司

Suzhou-based manufacturer of AI-powered wire-free robotic lawn mowers sold globally under the NexLawn brand, using 3D LiDAR and stereo vision.

Commercial Service Robots

HYFIX Spatial Intelligence

U.S. semiconductor startup building autonomous-systems GNSS chips and PNT modules for drones and robotics with centimeter-level RTK accuracy.

UFORCE

London defence firm formed 2025 from Ukrainian drone makers; builds MAGURA USVs, NEMESIS UAVs and LIUT UGVs, with 1,000+ MAGURA V7 fielded.

Trener Robotics

US-Norwegian Physical AI firm whose Acteris platform adds pre-trained, vendor-agnostic skills to ABB, Universal Robots and FANUC machine-tending cells.

Rhoda AI

Palo Alto robotics company building a video-prediction robot control model (DVA, FutureVision) and a wheeled 25 kg-payload robot; USD 450M Series A in 2026.

RobotToday Initiative

Robotics needs a service framework.

RSF defines a common language for robot service capability, lifecycle operations, certification pathways, and service-provider networks.

inJoin the RobotToday community on LinkedIn

Daily robotics news, in-depth analysis, conference highlights, and discussions with professionals worldwide.