The problem
Emergency teams, autonomous systems, and field operations can lose reliable satellite positioning exactly when conditions become most demanding.
Research
Robust AI for navigation when systems are under pressure.
A focused look at Theodore's KALMNav AI thesis and its contribution to GPS-denied emergency navigation, disaster response, autonomous systems, and adaptive AI.
Research thesis
KALMNav AI: emergency navigation when GPS goes down.
Theodore is developing an AI-enhanced emergency navigation system designed to maintain positioning accuracy during GPS downtime. The work contributes to disaster response, autonomous systems, and robust AI models that remain reliable in changing environments.
Signal stack
Sensor fusion for resilient positioning.
The framework integrates multiple signal sources, then uses deep learning to preserve useful navigation estimates when satellite positioning is unavailable.
IMUWiFiLiDARDeep learningGPS-denied navigation
Research intelligence
Designed for the moments where ordinary navigation becomes fragile.
Theodore's research is not framed as a laboratory novelty. It is a response to a practical failure mode: systems that depend on GPS can become vulnerable indoors, in dense urban settings, during disasters, or when infrastructure is disrupted.
The research direction
KALMNav AI combines IMU, WiFi, and LiDAR signals with deep learning so navigation estimates can remain useful during GPS downtime.
The larger ambition
Theodore's planned PhD direction extends this work toward lifelong learning and robust AI systems that adapt without becoming unsafe.
How the thinking develops
01
Gather imperfect signals
Treat each sensor as a partial witness rather than a perfect source, then design the system around uncertainty from the beginning.
02
Fuse and learn
Use statistical reasoning and deep learning to combine signal streams, identify useful patterns, and reduce drift in positioning.
03
Validate for pressure
Measure the system against the real question: can it support decisions when infrastructure, weather, buildings, or emergencies reduce GPS reliability?