On thirty years of
convergent preparation

People sometimes ask how the SciPhAI platform came to exist — a new class of LiDAR architecture, a decade-spanning mathematical framework for machine intelligence, two patent families filed years apart but pointing at the same thing. The honest answer is that it did not arrive suddenly. It was prepared for over thirty years, from three directions at once, until the directions converged and the answer became, in retrospect, obvious.

I want to explain that preparation — not as biography, but as argument. Because the preparation is the reason the platform is what it is, and understanding the preparation is the clearest way to understand why the technology works the way it does and why no one else has built it.

I completed a Master of Engineering and Ph.D. in Electrical Engineering at the University of New South Wales in Sydney, Australia, with a couple of theses covering various apsect of photonic devices and systems. The work that absorbed me most in those years was optical soliton communications — the propagation of self-sustaining light pulses through nonlinear optical fibers. At the same time never stopped working on figuring out how do we come to know something and why methematical formulations and modelings are so obyed by the nature and our universe. Theoretical modeling predict something, and if fundamental assumptions are not flawed, the nature will attest to that!

A soliton is a pulse that maintains its shape during propagation because the nonlinear response of the fiber exactly compensates for dispersive broadening. What fascinated me was not just the stability property but a related phenomenon: the soliton self-frequency shift. As a soliton propagates through a fiber, stimulated Raman scattering transfers energy from the high-frequency leading edge of the pulse to the low-frequency trailing edge. The carrier frequency of the pulse shifts continuously downward as it travels. The shift rate depends on the fiber's nonlinear coefficient, the dispersion, the Raman response time, and the pulse peak power — all knowable, all calculable.

I published in IEEE journals, Optics Letters, and Optics Communications across several distinct areas — photonic devices, fiber systems, nonlinear propagation, integrated optics. The range was deliberate. I was less interested in deepening one narrow channel than in understanding how the layers of photonics connected — device physics to guided propagation to system architecture.

In 1996 I founded Peleton Photonic Systems Inc. The idea was to build metropolitan-area optical networks delivering OC-48 — 2.5 gigabits per second — to residential subscribers. This was the year 2000 era. The ambition was real: multi-wavelength laser sources, external modulators, wavelength-division multiplexed transmission systems, all-optical routing. We raised US$22 million and built the technology.

What Peleton taught me was not primarily optical physics — I already had that. It taught me the discipline of making photonic technology perform reliably in the field, end to end, under real-world conditions. The gap between a laboratory demonstration and a deployed system is where most optical startups fail. Closing that gap requires a different kind of thinking than the physics requires — systems thinking, manufacturing thinking, the discipline of understanding failure modes before they occur.

In 1998, while Peleton was still operating, I co-founded Zenastra Photonics Inc. — a company focused on fabrication of silica-based passive and hybrid photonic integrated circuits. Zenastra raised US$44 million. The work was at the fabrication layer: planar lightwave circuits, waveguide splitters, arrayed waveguide gratings, hybrid integration of active components onto passive substrates.

Peleton operated until 2007. Zenastra until 2001. Between the two companies, I had built — as founder, CEO, and technical lead — one systems-level photonic company and one device-fabrication-level photonic company, in the same decade, at the same time that I was developing the mathematical framework that would eventually become the intelligence layer of the SciPhAI platform.

The question that occupied me in parallel with the photonics companies — and increasingly after them — was a different kind of question. Not how to transmit light. How does a system come to know something?

The answer I developed, beginning with a provisional patent filed in 2008 and a full application in 2009, was built on a single foundational data structure: the ordered participation matrix, PM^kl.

The participation matrix is not a simple co-occurrence count. It is an ordered, structured representation of how the components of any composition relate to one another across levels of granularity. In the 2008–2018 patent series I called these components "ontological subjects" — the entities of which any body of knowledge is composed, at any order of description. In the 2019 State Navigation specification I generalized the terminology to "state components" — because the framework applies not just to knowledge bodies but to any system navigating a state space.

Over the decade from 2009 to 2019, I filed and received approximately twenty issued patents in this framework, applying it to knowledge discovery, semantic processing, document analysis, genomic analysis, and artificial intelligence. The participation matrix appeared in each of these applications as the foundational structure. The specific form it took, the derivative objects computed from it, and the algorithms for navigating the knowledge space evolved — but the core object remained constant.

By 2019, two decades of photonics work and a decade of mathematical framework development had arrived at the same question from opposite directions.

From the mathematics side: the participation matrix framework could, in principle, process any body of data — text, genomic sequences, financial records, sensor streams. But what kind of sensory data would give a machine the richest possible representation of the physical world? A point cloud produced by a conventional LiDAR — N independent range measurements at N angles — is, informationally, a sparse and causally impoverished representation. It tells you distances. It does not tell you how those distances relate to one another, which objects caused which returns, or how the scene is structured causally. A participation matrix built from such data would be shallow.

From the photonics side: I knew, from two decades of building photonic systems, that the conventional LiDAR architecture — N independent laser-detector signal chains — was expensive, complex, fragile, and architecturally unchanged from its origins. And I knew, from the soliton years, that a propagating pulse in a nonlinear fiber naturally encodes its position in its own carrier frequency. That observation — dormant for twenty-five years — suddenly had a purpose.

The result is the Coded Propagation-Steered Illumination (CPSI) architecture — a guided medium with radiating apertures addressed by a single coded pulse packet. In its primary TDE mode — Temporal Direction Encoding — direction is identified from propagation delay alone, requiring only direct detection and no coherent receiver. In its secondary CFSRP mode, the carrier frequency of the pulse encodes direction simultaneously, adding velocity measurement and longer range. Both modes emerge from the same physical design. Neither requires external beam steering of any kind.

The State Navigation theory was filed in July 2019. It named LiDAR, cameras, and radar explicitly as sensory sources for the participation matrix. It introduced the Causal Association Strength Matrix — the asymmetric, directional measure that distinguishes what causes what, not merely what correlates with what. And it described a complete pipeline from physical sensing to rational, interpretable, explainable machine decision-making.

Five years later, in October 2024, the LiDAR hardware was filed — the Coded Propagation-Steered Illumination (CPSI) architecture, the CFSRP modality, the single omnidirectional detector, the participation matrix as native output. Not conceived after the intelligence theory. Motivated by it.

There is one more thing worth mentioning, because it speaks to the completeness of the framework and because it comes from the same cross-domain thinking that produced everything else.

In the State Navigation specification, I formulated the problem of man-machine conversation using coupled mode theory — the mathematical framework from photonics that describes how energy transfers between two waveguide modes when brought into proximity. In photonics, efficient energy transfer requires phase matching: the two modes must have compatible propagation constants for coupling to occur. Mismatched modes exchange little energy regardless of proximity.

The same mathematics, it turns out, describes knowledge transfer between a human and a machine in conversation. Efficient knowledge transfer requires that the machine's internal representational structure — its "mode" — be compatible with the human's knowledge structure. A machine that represents knowledge as a participation matrix with causal association measures can be phase-matched to the structured knowledge of a human expert. A machine that represents knowledge as statistical weights over tokens cannot, regardless of how many words it produces.

This formulation is in the patent record. It is one of the things that convinced me the framework was genuinely foundational — not a useful engineering tool, but a description of something real about the relationship between physical systems and knowing systems.

I want to be precise about what I am and am not claiming. The idea that intelligence is grounded in physical reality is not mine alone. Wiener alluded to it in 1948. Brooks circled it in 1986. Lakoff and Johnson named the phenomenon in language. Gibson described direct perception. They asked the right question.

What I am claiming is this: I gave an engineering answer. A specific, patented, physically realizable answer to the question those traditions raised but did not resolve. What must a sensor produce for a machine to genuinely understand its environment? What mathematical structure must the machine's knowledge take for its decisions to be rational, interpretable, and causally grounded? How does knowledge transfer efficiently between a human expert and a machine?

The answers are in the patent record, dated. The preparation that made those answers possible took thirty years. The vocabulary arrived in 2025. The work did not wait for it.