Skip to content

SCPN Control — The Fastest Open-Source Fusion Controller

SCPN Control

The Problem

Fusion energy is within reach, but real-time plasma control remains the bottleneck. Today's tokamak control systems are:

  • Slow — physics loops at 4--10 kHz, limited by Fortran/C legacy code
  • Fragile — no formal verification, no disruption prediction
  • Monolithic — tightly coupled to specific machines (DIII-D, ITER)
  • GPU-locked — P-EFIT needs CUDA for sub-ms reconstruction

scpn-control solves all four.

The Solution

A standalone neuro-symbolic control engine that compiles Stochastic Petri Nets into spiking neural network controllers — with formal verification, sub-ms latency, and zero GPU dependency.

11.9 microsecond control loop

Faster than any open-source fusion code. Faster than the DIII-D PCS physics loops. On commodity hardware. No FPGA. No InfiniBand.

Metric scpn-control DIII-D PCS TORAX ITER PCS
Control frequency 10--30 kHz 4--10 kHz Offline ~100 Hz
Step latency (P50) 11.9 us 100--250 us ~ms 5--10 ms
Language Rust + Python C/Fortran JAX TBD
GPU required No No Yes TBD

0.39 ms neural equilibrium — without GPU

P-EFIT achieves <1 ms on GPU hardware. scpn-control matches this on CPU only using a PCA + MLP surrogate trained on SPARC geometries.

Full-stack control in one package

Capability Status
Grad-Shafranov equilibrium solver Production
1.5D coupled transport Production
PID controller Production
H-infinity controller (Riccati) Production
MPC (gradient-based, surrogate dynamics) Production
Spiking neural network controller (Nengo) Production
Phase dynamics (Kuramoto/UPDE) Production
WebSocket live telemetry Production
Formal contract verification Production
ML disruption prediction (Transformer) Experimental
SPI ablation mitigation Experimental
Real-time digital twin Experimental
Neuro-cybernetic controller Experimental

Why It Matters

For Fusion Startups

You need real-time control now, not after a 3-year bespoke development cycle. scpn-control gives you:

  • Drop-in controller with 1969 tests and CI-gated RMSE validation
  • Runs on edge hardware (no data center required)
  • MIT/Apache-2.0 dual-licensed — no copyleft restrictions

For National Labs

Your PCS is decades old. scpn-control offers:

  • Modern Rust + Python stack replacing legacy Fortran
  • Formal verification via contract-based pre/post-condition checking
  • Digital twin for offline commissioning and operator training

For ITER / DEMO

The 100 Hz diagnostic cycle won't cut it for disruption mitigation. scpn-control's 30 kHz loop gives you the headroom for:

  • Real-time disruption prediction (ML-based, <1 ms inference)
  • SPI pellet injection with halo current / runaway electron physics
  • Closed-loop SNN feedback replacing hardwired interlock logic

Architecture

53 Python modules | 5 Rust crates | 1969 tests | 15 CI jobs

src/scpn_control/
+-- scpn/       Petri Net -> SNN compiler (formal contracts)
+-- core/       GS solver, transport, scaling laws
+-- control/    PID, MPC, H-inf, SNN, digital twin
+-- phase/      Paper 27 Kuramoto/UPDE engine (7 modules)

scpn-control-rs/
+-- control-types/    PlasmaState, EquilibriumConfig
+-- control-math/     LIF neurons, Boris pusher, Kuramoto
+-- control-core/     GS solver, transport, scaling
+-- control-control/  PID, MPC, H-inf, SNN
+-- control-python/   PyO3 bindings

Live Demo

Streamlit Dashboard: scpn-control.streamlit.app

Real-time 16-layer Kuramoto-Sakaguchi phase sync with global field driver. Interactive controls for coupling strength, oscillator count, and Psi driver.

Phase sync convergence (500 ticks, 16 layers x 50 oscillators):

Phase Sync Convergence

Validation

Every claim is CI-verified. Every benchmark is reproducible.

Validation Method Result
DIII-D shot replay 16 reference shots, RMSE gated < 15% Te RMSE
SPARC equilibrium 8 EFIT reference equilibria < 5% flux error
IPB98(y,2) scaling ITPA multi-machine database 26.6% RMSE
Kuramoto convergence R -> 0.92, V -> 0, lambda < 0 500-tick verified
Control latency Criterion benchmark (P50/P99) 11.9 / 23.9 us
Neural equilibrium PCA + MLP vs Picard ground truth 0.39 ms mean

Getting Started

pip install scpn-control          # From PyPI (v0.2.0)
scpn-control demo --steps 1000    # Closed-loop control demo
scpn-control benchmark            # PID vs SNN timing
scpn-control live --zeta 0.5      # Real-time WS phase sync

# Rust acceleration (optional)
cd scpn-control-rs
cargo test --workspace
cd crates/control-python && maturin develop --release

Publications

  • Paper 27: "The Knm Matrix" — 16-layer Kuramoto-Sakaguchi phase dynamics with exogenous global field driver. arXiv:2004.06344
  • Competitive Analysis: Full benchmark comparison against DIII-D PCS, TORAX, FUSE, GENE, JINTRAC, P-EFIT

Licensing
Open Source MIT OR Apache-2.0 — permissive, no copyleft
Contact protoscience@anulum.li
Organization ANULUM CH & LI
Authors Miroslav Sotek (ORCID), Michal Reiprich

Next Steps

  1. Try it: pip install scpn-control
  2. See benchmarks: Competitive Analysis
  3. Live demo: scpn-control.streamlit.app
  4. Talk to us: protoscience@anulum.li