Stanwell, the Queensland government-owned electricity generator, has deployed an artificial intelligence platform to improve how it charges and discharges batteries and forecasts power demand across the state's grid.
The Stanwell Modelling Platform (SMP) runs on Microsoft's Azure cloud infrastructure and uses AI services to process data and run predictive models across the company's trading, forecasting and safety operations.
Stanwell said battery optimisation had improved asset performance by more than 200%, while forecasting accuracy increased by 30% for certain time periods.
Scenario simulations, which help the company model different market and grid conditions, now run at up to 15 times the previous speed and at a fraction of the cost.
The platform began as a small-scale experiment on Stanwell's trading floor before expanding across the company.
Stanwell said the improvements help absorb excess solar generation during periods of peak output, reduce volatility in the wholesale electricity market and lower costs for customers.
The work is driven by the changing character of Queensland's grid, which is adding significant volumes of renewable generation and battery storage, creating new complexity for operators trying to balance supply and demand in real time.
Related reading
- OpenAI releases new faster and cheaper GPT-5.4 models
- Phone networks are being repurposed to carry artificial intelligence, Nvidia says
- PayPal launches its own digital dollar in 70 countries to speed up international payments
"Think of it as our intelligent command centre," said Kevin Lin, Stanwell's chief information officer, adding that the platform learns continuously as it takes in more data.
Stanwell said it plans to extend AI optimisation across its broader generation and storage portfolio.
The recap
- Stanwell deployed Stanwell Modelling Platform built on Microsoft Azure.
- Battery performance improved by more than 200 percent using AI.
- Company plans to extend AI optimisation across its generation portfolio.