Science Counter Inc filed patents on the theory before the market knew the category existed. Every implementation of that theory is covered โ not the specific product built, but the mathematical framework and physical architecture that make any product in the domain possible.
Before either foundational family was filed, eleven years of independent mathematical development produced the framework that both families rest on. These are not implementation patents โ they are framework patents covering the theory of causal association, participation probability, and rational navigation in systems that must reason under uncertainty.
The earliest prior art predates the AI filings by twelve years. Canadian patent CA 2186817, "Optical Communication System Utilizing Photonic Patterns," filed 1996, claims a system in which each communication channel is defined by a unique photonic pattern โ a participation signature of wavelengths. The mathematical structure is identical to PMkl: higher-order entities defined by the participation of their constituent elements, recognised by linear operations. Same inventor, two domains, one mathematical insight โ filed twelve years before the AI formalisation.
These substrate patents serve two functions. First, they establish the theoretical foundation with priority dates going back to 2008 โ making it impossible for a challenger to claim novelty on the core mathematics. Second, they create a web of prior art that any freedom-to-operate analysis for a competitor in causal AI must navigate.
All patents issued ยท Full list available under NDA
The complete participation matrix framework: PM^kl as the native data structure for causal association, Causal Association Strength Matrix (CASM), and rational state navigation under uncertainty. LiDAR, cameras, and radar named as primary sensory inputs.
The patent covers the mathematical framework, not a specific product. Any autonomous system that constructs a participation matrix from sensory data, calculates causal association strengths, and navigates states based on causal inference โ in any physical domain, using any combination of sensors โ falls within the scope of this family.
A competitor cannot design around this by using different hardware, different programming languages, or different sensor modalities. The theory is covered.
Jensen Huang named the category "Physical AI" at CES in January 2025 โ five and a half years after this patent was filed. Every company that entered the Physical AI market after July 2019, and every company working on causal sensing intelligence today, faces this prior art in their freedom-to-operate analysis. The market arrived at the theory. The theory was already protected.
Coded Propagation-Steered Illumination (CPSI), Temporal Direction Encoding (TDE) primary mode, Continuous Frequency-Shifting Radiating Pulse (CFSRP) secondary mode, single omnidirectional collector, PM^kl as native sensor output. CIP of PCT/CA2020/051000.
CPSI: Beam steering via propagation physics, not mechanics. One fiber coil replaces N independent channels. Cost does not scale with resolution.
CFSRP: Direction encoded in carrier frequency via nonlinear propagation. Zero commercial competitors. A different branch of the LiDAR technology tree, not an improvement on FMCW.
Windshield-embedded: The LiDAR is the glass. No external housing. A product category that does not currently exist commercially.
The State Navigation framework (Family 01) and the ISS Platform (Family 02) are available for licensing in domains where Science Counter Inc is not forming an operating company. This includes healthcare AI, genomics, legal intelligence, financial intelligence, and other verticals where causal reasoning over structured data is a core requirement. Licensing is handled through ATTVC, SCI's patent analytics and licensing division.
Begin a licensing conversation at attvc.com Read the public technology overview first Request technical edition (qualified parties) โ