MāyāPramāṇa

Universal Quantum Sensor Controller · 2026–present

Quantum Sensing C++23 Haskell Metrology

Valid Cognition of the Measured World

Every quantum sensor lab builds its controller from scratch — LabVIEW here, a custom FPGA there, Python scripts held together with hope and grad-student labour. Each controller works for one instrument, one configuration, one set of assumptions. The knowledge is trapped in the wiring, illegible, unreproducible.

MāyāPramāṇa is a different approach: not a piece of hardware but a grammar — a way of expressing what any quantum sensor needs. Digital twins of specific instruments, built one at a time, until the common structure emerges. The name comes from Indian epistemology: pramāṇa means valid cognition, the reliable path from raw perception to justified knowledge.

Architecture: Pure Core / Effectful Shell

Following the functional architecture established in MayaPortal, the system separates pure computation from side effects:

Three Languages, One Physics

LanguageRole
HaskellExecutable specification, QuickCheck properties — the physics as type-checked mathematics
C++23Deployment and performance, compile-time dimensional analysis, monadic composition from MayaPortal
PythonInteractive exploration, plotting, org-babel sessions, QuTiP integration

All three must agree on the same physics. Cross-language validation follows the MayaJiva pattern: Python generates reference data, Haskell and C++ load and compare.

The Universal Grammar of Measurement

The same epistemic structure appears across sensor types:

The same grammar describes an atomic magnetometer, an NV-diamond sensor, a gravimeter, an optical clock. The physics differs; the epistemology does not.

First Target: Atomic Magnetometer

The first digital twin models a Bell-Bloom atomic magnetometer: optical pumping, Larmor precession, RF driving, Faraday readout, lock-in demodulation, and PID feedback — each a pure function, composed into a controller, tested end-to-end against the Cramér-Rao bound.

Connections

Technical Stack

Haskell (QuickCheck, type-level physics) · C++23 (std::expected, concepts, constexpr) · Python (NumPy, SciPy, QuTiP, matplotlib) · Org-mode literate programming · Fisher information · Cramér-Rao bounds · Allan deviation

← Projects