An independent open-idea organization for difficult problems in science and engineering.
Austurusa studies problems that benefit from first-principles theory, computation, simulation, or physical experiment. Work is posted with its status made explicit; conjectures, research notes, simulations, measurements, implementations, and formal publications are not treated as interchangeable.
Can a continuous-time autonomous analog system be synthesized whose largest Lyapunov exponent follows a prescribed monotone transfer law over a robust chaotic interval?
Status: open, active · problem statement v0.2 · nonlinear circuits / dynamical systems
A workload-conditioned residual-graph framework for distinguishing GPU hardware deterioration from workload shifts and observability failures, with exposure-normalized health evidence and active diagnostic selection.
Status: Draft v0.1 · research note · not peer reviewed · full paper [PDF]
Battery-powered Geiger-Müller radiation counter integrating high-voltage generation, embedded pulse acquisition, LCD readout, custom PCB hardware, and experimental calibration against a commercial reference instrument.
Status: completed hardware experiment · 2022 · full report [PDF]
A reference implementation for permissionless settlement of off-chain signed payment intents on Tempo, with replay, expiry, cancellation, wallet, and multi-relayer failure handling.
Status: reference-grade implementation · demo-validated on Tempo testnet · not production-audited
Problems are stated precisely enough to fail. The normal sequence is prior-art search, derivation, numerical validation, physical or computational experiment where appropriate, and publication. Negative results are retained when they eliminate a plausible architecture or sharpen the problem.