Controlling the nation's electrical power grid promises to grow insanely complex. Renewable energy sources are being added, more electric vehicles are plugging in, aging technology is being upgraded and tons of new data will flow. To deal with all that and increase efficiency, engineers are tapping directly into brain cells.
Brain in a Dish Comes Alive
This new approach to power management is being led by Kumar Venayagamoorthy, a professor of electrical and computer engineering at Clemson University. In thinking about ways to optimize the electrical power grid, Venayagamoorthy looked for a system that could monitor, forecast, plan, learn, make decisions, he told LiveScience.com. What does that better than the human brain?
He got help from Georgia Institute of Technology neuroengineer Steve Potter, who developed a method for growing neurons in a dish containing electrodes to better understand how the brain responds to information. Venayagamoorthy came up with a power grid computer simulation and together the scientists connected it to living neurons from rodents. This system allowed the scientists to stimulate and record activity from the cells.
Electric Vehicles Won't Bring Down the Power Grid
Then the engineers successfully "taught" the living brain cell network how to respond to complex data and incorporated those results into a bio-inspired artificial network, according to LiveScience.com. The overall project, dubbed Brain2Grid, ultimately aims to come up with a smart brain-like control system for the power grid.
While the scientists are using living brain cells for their research, they say they don't plan to have them hooked up forever. Instead, they want to emerge with a super-intelligent computer program that can handle the electrical grid of the future. Forget smart. The next generation electrical grid could be sentient.
Photo: Microscope image shows fluorescent-labeled neurites extending from dorsal root ganglia growing on a thin polymer film made of aligned nanometer-diameter fibers. Credit: Isaac Clements