Quantum Synthetic Data Generation for Industrial Bioprocess Monitoring

TL;DR

The paper tackles data scarcity in industrial bioprocess monitoring by introducing a Quantum Wasserstein GAN with Gradient Penalty (QWGAN-GP) that uses a Parameterized Quantum Circuit (PQC) as the generator to synthesize high-fidelity time-series data, validated on Optical Density measurements. It presents an eight-stage, decision-driven framework integrating sensor feasibility, mechanistic and data-driven modeling, and quantum data generation, with a CNN critic and rolling-window preprocessing. Empirical results on a 20 L photobioreactor dataset show strong distributional fidelity (QQ plots, PDF/CDF fidelity) and temporal structure preservation, with a DTW score of 0.6843 indicating improved temporal alignment relative to prior methods, supporting improved soft-sensor development and predictive control under data limitations. The work contributes to open science through public code and data, discusses practical challenges such as quantum hardware noise and small datasets, and outlines future directions including Hybrid-GAN architectures, error mitigation, and scaling to multivariate bioprocess data, highlighting the potential impact on robust biomanufacturing and process optimization.

Abstract

Data scarcity and sparsity in bio-manufacturing poses challenges for accurate model development, process monitoring, and optimization. We aim to replicate and capture the complex dynamics of industrial bioprocesses by proposing the use of a Quantum Wasserstein Generative Adversarial Network with Gradient Penalty (QWGAN-GP) to generate synthetic time series data for industrially relevant processes. The generator within our GAN is comprised of a Parameterized Quantum Circuit (PQC). This methodology offers potential advantages in process monitoring, modeling, forecasting, and optimization, enabling more efficient bioprocess management by reducing the dependence on scarce experimental data. Our results demonstrate acceptable performance in capturing the temporal dynamics of real bioprocess data. We focus on Optical Density, a key measurement for Dry Biomass estimation. The data generated showed high fidelity to the actual historical experimental data. This intersection of quantum computing and machine learning has opened new frontiers in data analysis and generation, particularly in computationally intensive fields, for use cases such as increasing prediction accuracy for soft sensor design or for use in predictive control.

Figures (13)

• Figure 1: Conceptual diagram of how we envision using quantum generative AI within the industrial bioprocess
• Figure 2: Illustration of classical GAN architecture
• Figure 3: Classical WGAN-GP architecture
• Figure 4: Quantum circuit used as generator
• Figure 5: Decision-tree of the bioprocesses monitoring framework