ElectroAI: AI-driven electrochemical system optimisation

Electrochemical CO2 conversion is central to sustainable chemical manufacturing, yet practical deployment remains constrained by inefficient and unstable operation. Current electrolysers rely on fixed settings that cannot adapt to dynamic changes in device behaviour, resulting in inefficiencies and rapid degradation.

ElectroAI aims to establish an AI-assisted control framework for electrochemical energy systems that autonomously optimises performance and stability. The project aligns AI decision-making with underlying physical and electrochemical principles, enabling scalable, efficient and safe operation of next-generation CO2 electrolysers for clean energy and sustainable chemical manufacturing.

The collaboration with MDAP aims to embed data-driven intelligence into the control framework by co-developing reinforcement learning actor–critic algorithms and time-series data infrastructure.

To do this, the team will leverage MDAP’s expertise in machine learning, data engineering and algorithm optimisation to:

• Develop efficient RL architectures for real-time electrolysis control
• Integrate constraints and electrochemical models into AI decision loops to ensure scientific validity
• Establish best-practice workflows for managing experimental datasets
• Demonstrate the framework using data from sensor-enriched electrolysis platforms.

This project will deliver new methodologies for coupling AI with electrochemical models and open-source datasets for analysing electrochemical systems. The resulting framework will accelerate the transition from empirical testing to autonomous, data-driven operation, generating insights into device behaviour and enabling predictive operation.

By advancing intelligent control technologies for hydrogen production, CO2 utilisation and other clean-energy systems, this research will improve operational efficiency and lifespan of electrolysers, reducing waste and maintenance costs. These outcomes will strengthen Australia's capability in renewable manufacturing and emissions reduction technologies, contributing to the global net-zero transformation.