About FSME Logic

Built on a
question about
the universe.

FSME Logic was built by an independent researcher in Edmonton, Alberta — working alone, on consumer hardware, driven by curiosity about how physical systems actually behave over time.

It started with a question about the universe. Alexander Kalyniuk — founder of FSME Logic — began by watching physics documentaries late at night, trying to understand how the physical world actually works at a fundamental level. That curiosity became an obsession, and the obsession became a research project.

Working independently, with no institutional affiliation and no external funding, Alexander began formalizing a theoretical framework to describe something he kept observing across different physical systems: The universe is unique. And if you understand that uniqueness, you can map systems degradation.

What started as theoretical physics became something unexpected. The same structural signatures that appeared in astrophysical data showed up in industrial sensor telemetry, in commercial vehicle fleet data, in bearing vibration records. The detection principle was domain-independent. If you could read the history of a signal, you could predict its failure.

The theory was formalized and filed for patent protection before a single commercial dataset was tested. The theory preceded the measurement — the parameters were not fitted to data after the fact. When the first validation results came back on NASA and ESA datasets, they confirmed what the theory had already predicted.

FSME Logic is the commercial application of that research. Fully offline. Air-gapped. Deployable on edge hardware. No cloud. No new sensors. No cloud processing. No third-party platforms.

The long-term vision is larger than predictive maintenance. The same detection principle that identifies mechanical failure months before it happens has implications across many physical domains. FSME Logic is the first commercial implementation of that principle. It won't be the last.

What kept Alexander going through the hard parts was straightforward: the knowledge that building something real could build a better future for his family. That's still the reason.

// Location

Edmonton, Alberta, Canada
Operating independently since 2025

// Research Basis

Independent theoretical physics research — no institutional affiliation, no external funding. Consumer hardware. Open datasets. Reproducible results.

// Theory Before Data

The detection framework was formalized before commercial validation datasets were tested. The theory preceded the measurement.

// Validation Record

6+

Independent public datasets validated — NASA, ESA, CWRU, IBM, commercial refrigeration, commercial trucking fleets.

// IP Status

Detection methodology protected as trade secret. Core algorithms and calibration constants never leave air-gapped hardware.

// Cyber Security

Architecture aligns with NIST SP 800-171 and the forthcoming Canadian Program for Cyber Security Certification (CPCSC). Level 1 self-assessment in progress.

"Standard monitoring assumes the universe has amnesia.
The physics says it does not.
FSME Logic reads the memory."
— Alexander Kalyniuk, Founder, FSME Logic

Independent validation.
Real datasets. Real results.

Every validation result was produced on offline edge hardware against publicly available datasets — the same datasets used to benchmark the world's leading predictive maintenance algorithms.

🛰️

Blind Testing

The ESA OPS-SAT validation was conducted with no prior access to the dataset. Results were compared against official ESA ground-truth labels after the fact. No retrofitting. No parameter tuning after seeing the data.

⚙️

Edge Hardware Only

Every result was produced on ARM-based edge hardware — Raspberry Pi processors, offline laptops. No cloud compute. No GPU clusters. The same hardware used in field deployments.

📐

Zero Training Required

Unlike competing algorithms that require extensive historical failure libraries, FSME Logic was deployed cold on every dataset — no prior examples of failure provided before testing.

🔬

Reproducible Methodology

The detection methodology is documented in sufficient detail for independent replication. Public datasets. Documented parameters. Results that can be independently verified.

Built to last.
Built to matter.

FSME Logic began as a maintenance detection engine for industrial and commercial fleets. That remains the immediate focus — helping Edmonton's industrial operators catch equipment failure before it becomes catastrophic, using technology validated against the most demanding datasets in aerospace and industrial engineering.

But the detection principle has a longer arc. The same physics that identifies mechanical failure in a truck fleet months before breakdown has implications for many other domains. The goal is to put this detection capability inside the hardware that manages critical systems — to build something small enough to run anywhere and powerful enough to matter everywhere it runs.

That's the long game. Right now, the next step is simpler: proving the value of this technology, one audit at a time, with clients who need it.

Ready to see what
FSME Logic can find?

Request a Pilot Integrity Audit — send us your existing sensor data and receive a comprehensive forensic PDF report. No new hardware required.

Contact Us ▸ View Validation Data