Optimising the energy system as it gets harder to schedule

The problem

Every power system solves the same problem repeatedly: which generators and storage to run, and at what level, to meet demand at lowest cost within a web of physical and network constraints. It is NP-hard. As renewables and storage grow, the problem gets bigger, conditions shift faster, and operators must solve it more often and further ahead. The tools that schedule the grid are becoming the limit on how well it can run.

Where we think we can add value

In the United States, grid-operator day-ahead scheduling runs to hundreds of thousands of variables against a roughly twenty-minute deadline; operators stop at an optimality gap because proving the best answer in time is not feasible. Closing that gap and extending the look-ahead is where we believe we can help. In Australia, the National Electricity Market dispatches every five minutes with prices swinging from minus

Status

What is measured and what is not. The market figures on this page come from public research and are established context. Our measured results to date are in mine planning. The performance we describe for energy is a hypothesis we want to test on a real problem with a real operator, not a delivered result. See /mining.

We have solved a problem of this class at scale: a scheduling problem of more than 200,000 variables returned in a few minutes on a single consumer-grade GPU, against legacy tooling that takes hours or days.