Overview
The problem, what the approach does, and where it stands, stated plainly.
Applied research
Software that makes safe decisions from data it cannot fully trust.
Each research program starts from a specific operational problem and produces a working reference implementation. The results below are from simulation, and each page states what has not been shown.
Project SENTINEL
Deterministic integrity monitoring for multi-source navigation. It checks sources against each other, flags the one that has been jammed, spoofed or has drifted, refuses to act when too few sources agree, and records every decision in a form that can be replayed.
Four questions
A four-part series on how a multi-source system quietly stops being trustworthy, one part per application below.
By sector
Maritime and defense · Aviation and autonomous systems · Critical infrastructure · Cybersecurity and AI agents
Patent portfolio
The inventor, Rajan Vasudevan, has filed the U.S. provisional patent applications below. They are patent pending. A provisional application is not examined, and none of these has been granted. A provisional application lapses unless a later application is filed.
SENTINEL
Deterministic multi-source consensus and accountable ingestion
- Provisional application
64/144,562 - Provisional application
64/149,078 - Provisional application
64/159,476 - Provisional application
64/161,448
Computer Vision
Consent-constrained media alteration and cross-modal verification
- Provisional application
64/179,643