New Delhi: Over a decade ago, NASA began working on an AI vision technology for its interplanetary spacecraft, such as its rovers and drones on Mars. Through a Small Business Technology Transfer programme, NASA funded the development of a neural network software for artificial intelligence, based on the human brain. The idea was to enable vehicles on extraterrestrial surfaces to autonomously navigate the terrain, by recognising features in the environment in realtime. An artificial intelligence company from Boston, Neurala, adapted the technology for drone navigation and collision avoidance. In both the applications, the software could run on the vehicle itself, with the limited resources available to an interplanetary surface craft.
Neurala software being used to evaluate quality of prepared meals. (Image Credit: Neurala Inc).
The design required that the decisions be made by the device itself, without relaying signals to a decision maker on Earth. The features of the technology that made it suitable for use in a deep space environment, also made it suitable for use on assembly lines on Earth. Neurala has adapted the software to monitor assembly lines in factory floors, which may contain hundreds of items passing through every minute, making visual inspections by humans challenging. The same AI technology that allows Martian vehicles to evaluate the environment in realtime, is also now used to monitor the assembly lines on Earth.
NASA spinoffs benefits Earth
Cofounder of Neurala and its CEO, Massimiliano Versace says, “Our software can learn very quickly on a processor with a very small footprint, a skill we learned working with NASA. By doing so, we enable vision inspection with whatever components are already available, deploying in minutes. In our exploration of the market, we realized that the manufacturing space had a precise need for this technology.” Neurala’s AI software is another example of technology originally developed for applications in space, benefiting humans living on Earth.
