Neurala to turn robots into ‘adaptive, learning beings’
June 27, 2013
Cambridge startup Neurala expects that its software for giving intelligence and autonomy to robots will part of commercial products by the end of the year, CEO and co-founder Max Versace said in an interview with Boston Business Journal.
The software aims to allow humans to control robots by telling them what to do, instead of operating them using remote control or programming for a specific task.
Neurala says its software aims to serve as the “brain” for robots, allowing them to process information sources and learn in a similar way to the human brain — ultimately enabling robots to work autonomously from humans.
Frst commercial application: toys, followed by applications in the enterprise in 2014.
Neurala’s two paying customers so far are NASA and the U.S. Air Force, which have paid the company $1.7 million since 2012 to develop software for a Mars Rover-like vehicle which could operate without needing to be guided, he said.
How it works
Neurala’s software works by taking in information from multiple sensors, processing the information sources simultaneously (as the human brain does) and then making decisions based on the information.
“It will allow the robot to learn a map of the environment, and allow the user to tell the robot, ‘Go find Bob, go to the conference room, go find this object,’” Versace said. “The user can tell this to 100 robots at the same time. So the robots become a multiplier of a person’s ability to work.”
Neurala currently has six partners working with the company — among them iRobot and Teledyne — which could eventually become customers for the technology, he said.