This Robot Knows About Itself, and No Human Helped it Realize

Scientists reveal building a robot that can understand itself, without any human intervention

The robotics industry is advancing at a rapid pace. Robots, powered by artificial intelligence, are making their way through every global industry, transforming the core of their operational strategies. Researchers from the Columbia University School of Engineering and Applied Sciences have built the first-ever robot that can learn about its entire self, from scratch, without human intervention! They show how the robot has constructed a kinematic model of itself and then leveraged that model to plan movements, attain goals, avoid obstacles, and then identify and repair body damages in a range of settings.

How did they make it work?

Firstly, a robotic arm was placed inside a circle of five streaming video cameras, the robot was then able to watch itself through the cameras while it swung freely. It moved back and forth to monitor exactly how its body moved in response to different motor commands, after three long hours, it finally stopped. The machine's internal neural network algorithm finished learning the relationship between the robot's motor action and the occupied volume around itself.

After this research method, the robot's consciousness of self-image began to emerge, which was depicted through a graph that the bot itself generated to show where it predicted its arm to be when it moved. As per reports revealed by the researchers, the robot's self-model was accurate to about 1% of its workspace. This endeavor comes as a part of Lipson's long-term goal of giving robots self-awareness.

AI consciousness has come a long way since Google's LaMDA raged the tech industry with debates about the AI chatbot being 'sentient'. Researchers have accepted that they have made a significant proportion of advancements in terms of tech consciousness. This initiative is just another step towards attaining consciousness in machines.