Atlas’ feet managed to cross the uneven surface of wood paths without encountering any difficulties.
If you are not passionate about artificial intelligence, you probably won’t care that Google’s Atlas humanoid robot took his first walk in the woods on Monday. For the rest of the world, however, this is incredibly good news.
Sure, the humanoid bot may look like a dancing monkey or a helpless old man trying to make his way to the nearest toilet in the night, but scientists have great expectations, nonetheless. After months of indoor experiments, the robot was finally tested on the uneven surface of the woods and it did remarkably well.
For the current test, MIT researchers, who are working with Boston Dynamics and Google for their new prototype, have used a power tether to ensure the stability of the robot. The video that was published on their official YouTube channel shows the bot tilting from one leg to another, yet maintain its balance on the hard wood paths.
The tasks grew gradually difficult as researchers also programmed the machinery to climb on small hills. The machine remained stable at all times and no problems have been encountered during the experiments.
MIT engineers will use the recent material to determine what improvements should be operated on the robot in order to become more flexible and, as a consequence, more reliable. So far, engineers have agreed that they will have to work on the mobility of the robot.
Previous indoor tests have helped scientists perform various upgrades on Atlas at the beginning of the year. Based on their description, the robot has been foreseen with lighter materials and improved actuators, so it can easily stand up after falls.
The humanoid machine is just as functional as a human being, mainly due to its flexible wrists. The bot can grab doorknobs and rotate them without using its entire arm – a very important feature for artificial intelligence prototypes.
Researchers plan to perfect artificial intelligence to the point where it can completely replace human beings. By doing so, robots can be programmed to carry out dangerous tasks and missions that are currently operated by human workers.
As surreal as this scenario may seem, engineers think scientists will soon be able to use robots to rescue people from disaster areas or to defuse bombs. For that, machineries will have to be endowed with sufficient independence to be able to move around and act without additional help.
Image source: www.dailymail.co.uk