About
I am an independent researcher working on periodic geometry, exact inequalities, and networks across delay and distance. The current focus is mathematics: the spaces between packed spheres, the passages through them, and the sharp constants that control positivity. The network work asks how information moves when links are intermittent and propagation is slow.
The through-line is a single end — getting humanity to space, and keeping the signal alive once we are there. Sustained operations behind the Moon need dependable relay infrastructure. Getting there is one problem; keeping the message moving is another.
I work without an institutional affiliation. That makes verifiability especially important: proofs, preregistrations, public negative results, archived artifacts, and permanent identifiers let the work be judged independently of its source.
Method
The standard has to be mechanical, and honestly a little unglamorous. Here is what it comes down to.
The mathematics papers give arguments to read and inspect, with exact replay code for the computational steps they use. In verified-sabr, the stated routing theorems are also formalized in Lean 4. A human proof, a finite computation, and a machine-checked theorem are identified separately, so you can tell what was established and how.
Before I examine the results, I write down the hypothesis, the thresholds, and the conditions that would prove me wrong, and freeze it. So when a prediction misses, you read the miss in the same words it was registered in. The goalposts are set down and timestamped before the game starts.
The TIN record includes useful routing results, reproducible measurements, and corrections to earlier interpretations. A delivery-ratio identity was once given more explanatory weight than it could bear; broader universality claims did not survive testing. Those corrections belong alongside the results that remain supported. The papers state their scope, and the public code remains available.
Before a result leaves here, it goes to review passes whose whole job is to try to break it. The published record states what survived, what failed, and what remains unresolved. The method for doing that is written up and public too — you are welcome to point it at this very page.
Papers ship with the code, the frozen inputs, and content hashes; data and software get permanent DOIs. Wherever a number turns up in a sentence, a named, rerunnable artifact stands behind it. No number should have to ask for your faith.
The record
Here is what that standard produces in practice — each piece public and checkable today.
verified-sabr formalizes Schedule-Aware Bundle Routing in Lean 4 and tests a reference implementation against it across thousands of queries. A deposited regime note (concept DOI 10.5281/zenodo.21316172) states what the standard pins down and what it defers. The differential testing turned up concrete discrepancies, which I filed as public issues against NASA's open-source HDTN. Whether anyone upstream picks them up is out of my hands, but the reports are on the record.
The mathematics papers, delivery-assurance and classification work, software, and ground-segment datasets are archived on Zenodo under a stable identity, each with its own DOI. None of it evaporates if this site does.
The way this research operates — claim classes, honest ledgers, refute-before-ship — is written up as its own public methodology. The process is open to inspection, not only its results.
Contact
Corrections and hard questions are especially useful — they help independent work stay honest.
This is independent research and is not affiliated with or endorsed by NASA or any other institution. Project names and issue reports identify the systems studied; they do not imply collaboration or endorsement. Some material documents work in progress and is identified as such. Corrections are welcome.