Bykov’s and Corzo’s technology, named GreenSight, offers AI-based solutions for reducing energy consumption and increasing data security. Their research, funded by DOE Advanced Scientific Computing Research program, provides seamless integration of software services while championing sustainability and ensuring data privacy throughout each phase of processing. Their technology is an energy-optimized AI module, designed to seamlessly integrate with conventional hardware, removing the necessity for additional computing resources.The technology is AI-based, but unlike standard AI tools, it can be used locally, on a user’s personal laptop, for example.

“That was part of the value of it,” Corzo said. “Training an AI model requires a lot of energy. We’re trying to tackle security with something that can run locally, without specialized hardware, and make it energy efficient. We developed this model with the idea of how the human brain works.”

For example, when reading a book, the brain does not save all the data, but the reader retains certain important pieces of information, Corzo explained. Standard AI saves all the information in the book, which requires a lot of energy and resources. But GreenSight uses less energy and can do 10 times more tasks using only a fraction of the energy, while providing the same insight. Thus its name.

The other team members in ORNL’s Cohort 18 group, Tobin, senior R&D staff and principal investigator for the team, and Kumar, R&D staff member, learned how to commercialize a technology that could help the wind energy industry meet national goals for carbon reduction.

In research funded by DOE’s Advanced Materials and Manufacturing Technologies Office and Wind Energy Technologies Office, Kumar developed IndraGuard, a nano-engineered multi-functional coating to enhance the lifespan of wind turbine blades and reduce their service downtime by improved lightning-strike protection. Lightning strikes and extreme weather account for almost 80% of downtime for wind turbines, causing high repair and logistics costs. In one year, more than 77,000 lightning strikes occurred on wind turbine blades in the U.S., with each turbine getting struck by lightning once or twice a year, on average.

This Oak Ridge National Laboratory news article "Energy I-Corps steeps scientists in the world of business, commercialization" was originally found on https://www.ornl.gov/news