quantum-safe-control-systems

name	quantum-safe-control-systems
category	systems-engineering
description	Safe deployment patterns for quantum control policies in cyber-physical systems. Covers Q-DASC discrepancy-attributed safe quantum control framework, Simplex architecture for quantum-classical switching, and certified safety layers for variational quantum circuit controllers. Based on arXiv:2606.28834 (Q-DASC) and arXiv:2606.31056 (Simplex Q-CPS).
trigger_words	["quantum control safety","safe quantum control","quantum policy deployment","variational quantum control","Q-DASC","simplex quantum architecture","quantum cyber-physical systems","QA-HSGPR","certified quantum safety","quantum model misspecification","quantum HVAC control"]

Quantum-Safe Control Systems

Safe deployment patterns for quantum control policies in cyber-physical systems, integrating certified classical safety layers with variational quantum circuit (VQC) controllers.

Core Papers

• Q-DASC (arXiv:2606.28834): Discrepancy-Attributed Safe Quantum Control — wraps VQC policies with certified classical safety layers
• Simplex Q-CPS (arXiv:2606.31056): Simplex-inspired architecture integrating quantum-assisted GPR with classical GPR for CPS

Q-DASC Framework

Architecture

┌─────────────────────────────────────────────────────────────┐
│                    Q-DASC Safety Wrapper                     │
├─────────────────────────────────────────────────────────────┤
│  ┌──────────────┐    ┌──────────────┐    ┌───────────────┐  │
│  │ VQC Policy   │───▶│ Discrepancy  │───▶│ Classical     │  │
│  │ (Quantum)    │    │ Detection    │    │ Safety Layer  │  │
│  └──────────────┘    └──────────────┘    └───────┬───────┘  │
│                                                  │          │
│                                    ┌─────────────▼───────┐  │
│                                    │ Violation           │  │
│                                    │ Attribution Engine  │  │
│                                    │ - Policy error      │  │
│                                    │ - Model error       │  │
│                                    │ - Physical limits   │  │
│                                    └─────────────────────┘  │
└─────────────────────────────────────────────────────────────┘

Key Components

1. Discrepancy Detection: Uses false-discovery-rate (FDR) control to identify operating regimes where the thermal model is misspecified
2. Model Repair: Applies shrinkage to repair local thermal gains in misspecified regions
3. Safety Projection: Projects the quantum policy output onto the repaired comfort-feasible set
4. Violation Attribution: Classifies residual violations into policy error, model error, or physical limits

Performance Results

Metric	Raw VQC	Model-Trusting	Q-DASC
Comfort violation	26.0%	55.3%	0.02%
Under NISQ noise	—	—	0.24%
Repair-aware variant	—	—	0.00%

Simplex Architecture for Quantum CPS

Design Pattern

                    ┌──────────────────┐
                    │  Runtime Monitor  │
                    │  (Safety Check)   │
                    └────────┬─────────┘
                             │
              ┌──────────────┼──────────────┐
              │              │              │
    ┌─────────▼─────┐  ┌────▼────────┐    │
    │ High-Perform.  │  │ High-Assur. │    │
    │ QA-HSGPR       │  │ Classical   │    │
    │ (Quantum)      │  │ GPR         │    │
    └───────────────┘  └─────────────┘    │
              │              │             │
              └──────┬───────┘             │
                     │                     │
              ┌──────▼───────┐             │
              │  Output to   │             │
              │  CPS Plant   │             │
              └──────────────┘             │

Principles

1. Dual-Module Design: High-performance quantum module + high-assurance classical module
2. Runtime Monitoring: Continuous safety evaluation during operation
3. Dynamic Switching: Automatic failover based on safety certification
4. Controllable Trade-off: Tunable balance between performance and safety assurance

Implementation Steps

Step 1: Train VQC Policy

• Use variational quantum circuits for the control policy
• Optimize on nominal model data
• Validate on held-out scenarios

Step 2: Build Discrepancy Detector

• Collect residuals between model predictions and observations
• Apply FDR control to identify statistically significant discrepancies
• Map discrepancy regions in the operating space

Step 3: Model Repair

• Apply shrinkage estimation to local model parameters
• Use historical data to repair misspecified thermal gains
• Maintain repair history for attribution

Step 4: Safety Projection

• Define comfort-feasible set based on physical constraints
• Project VQC policy output onto feasible set
• Measure projection distance as safety margin

Step 5: Attribution Engine

• Track violations after projection
• Classify into: policy error (VQC suboptimal), model error (unrepairable), or physical limits (infeasible)
• Use classification to guide system improvement

Pitfalls

• NISQ Noise: Finite-shot and depolarizing read-out noise can corrupt quantum policy outputs; the safety layer's classical projection is noise-invariant
• Model Trusting: Never deploy VQC policies without safety wrappers — raw VQC shows 26% comfort violation
• Over-Repair: Excessive shrinkage can over-smooth legitimate dynamics; use FDR to control false positives
• Energy Trade-off: Safety projection may increase energy consumption; optimize repair-aware VQC to minimize interventions

When to Use

• Deploying quantum reinforcement learning policies in safety-critical CPS
• Building energy management (HVAC) with quantum control
• Any scenario where model misspecification can cause safety violations
• NISQ-era quantum policy deployment where read-out noise is significant

References

• arXiv:2606.28834 — Q-DASC: State-of-the-Art Safe Quantum Control for HVAC under Local Model Misspecification
• arXiv:2606.31056 — A Simplex-Inspired Architecture for Integrating Quantum Capabilities into Cyber-Physical Systems
• arXiv:2606.31321 — Projection Operator Stochastic Equations for Non-Markovian Quantum Systems Under Continuous Measurement-Based Feedback