Qubit Engineering debuts AI-powered tool to analyze the modern power grid

Qubit Engineering, part of the 2023 Innovation Crossroads cohort at the Department of Energy’s Oak Ridge National Laboratory, recently launched Neuro-Grid, a computational tool that uses artificial intelligence for real-time, high-accuracy analysis of the power grid to offer insights not available through classical computing. Qubit leveraged Frontier, the world’s first exascale supercomputer housed at ORNL’s Oak Ridge Leadership Computing Facility (OLCF), to develop Neuro-Grid.

What’s the technology

• As a graphical neural network, Neuro-Grid is designed to meet the needs of a modern, evolving energy system by analyzing raw grid data and delivering insights quickly and efficiently, even with limited or missing data.
• Qubit is developing a component that will provide contingency and risk analysis, which could evaluate different scenarios, such as weather events or localized demands on the system. This information is critical to maintaining a reliable and resilient energy grid.

What’s the impact

• Qubit is submitting proposals to local power companies within the Tennessee valley, which comprises seven states, to cater Neuro-Grid for their usage, and map and analyze their distribution system.

ORNL’s Innovation Crossroads Advantage

• With support from the Tennessee Valley Authority (TVA) and the Department of Energy Office of Science’s Advanced Scientific Computing Research program, Qubit advanced R&D through Innovation Crossroads and partnered with Oak Ridge National Laboratory and TVA for critical testing and validation.
• Qubit leveraged Frontier, the world’s first exascale supercomputer housed at ORNL’s Oak Ridge Leadership Computing Facility (OLCF), to develop Neuro-Grid. With the resources of OLCF, Qubit was able to optimize Neuro-Grid to run on the cloud, making the software accessible for a variety of industry partners without losing computational power.
  • “Using Frontier helped us understand the model and its limitations. By rethinking our model, we downsized the computational resources needed. We could make the model more efficient, requiring less data and fewer computational resources,” said Marouane Salhi, Qubit CEO and co-founder.
  • “ORNL was the testbed for our ideas, and it set us on our way for further development. Through Innovation Crossroads, we had access to experts who have years of expertise in different fields,” Salhi said. Salhi was an ORNL postdoctoral research fellow from 2016 to 2018.
• The OLCF is a DOE Office of Science user facility and part of the Genesis Mission, DOE's initiative to secure American leadership in AI for science, energy, and national security.