Why does 'video of a humanoid robot making coffee' represent a big step for the future of AI robotics?
The ``video of a humanoid making coffee'' released by the American robotics company
Footage of Figure 01 making coffee could signal a huge leap in the future of AI robotics | Live Science
https://www.livescience.com/technology/robotics/video-of-figure-01-robot-making-coffee-could-signal-a-huge-step-in-the-future-of-ai-robotics
Figure is a robotics company based in California, USA, and in January 2024, it announced that it had signed a commercial agreement with car manufacturer BMW to supply humanoid robots for car production. The ``video of a humanoid making coffee'' released by Figure on YouTube on January 7 is attracting attention as it shows important progress in AI robotics.
The video actually released by Figure is below.
Figure Status Update - AI Trained Coffee Demo - YouTube
A test was conducted on January 5, 2024, in which a humanoid named ``FIGURE 01'' was made to make coffee using a coffee machine. FIGURE 01 was trained by watching 10 hours of footage.
The man places the coffee machine and capsule on the table in front of FIGURE 01, and sets the mug. Then say, 'Hey, FIGURE 01. Can you make me a cup of coffee?'
FIGURE 01 opens the lid of the coffee machine, pinches the capsule and sets it.
Close the lid and press the button.
FIGURE 01 patiently waiting for coffee to be poured.
When the coffee was poured, the man took the mug and thanked FIGURE 01.
At first glance, it looks like FIGURE 01, which simply learned how to make coffee using a coffee machine, simply made coffee according to a set procedure. However, it is explained that FIGURE 01 improved its operation by learning and correcting mistakes made in previous tests such as ``the capsule could not be set properly.''
For most people, using a coffee machine to make coffee is easy, but it requires a lot of motor skills such as opening the lid of the coffee machine, pressing a button, and lifting the cup without spilling the coffee. This is a complex task for robots to learn and perform, as it requires precise manipulation and correct steps. Therefore, the fact that FIGURE 01 learned how to make coffee just by watching the video and further fine-tuned its movements to correct mistakes is a major advance for AI robots.
'The robot's eyes don't just see what's happening in the coffee-making process. They don't just observe, they analyze the process and make sure everything is as accurate as possible,' said AI entrepreneur
Claire Walsh , who teaches data science and data ethics at the Open Data Institute (ODI), a British non-profit organization, said that FIGURE 01 was automatic in that the training was based on probability rather than basic rules. It is pointed out that it has something in common with the driving car. ``The number of movements in FIGURE 01 is impressive, and its skill and ability to self-correct may foreshadow future developments in this field,'' he said, adding that with appropriate training data, They believe that AI robots will be able to expand and diversify their movements.
On the other hand, AI expert Mona Kirstein said that FIGURE 01 is not a market-ready product, but rather a great first step in the industry. “Achieving human-level flexibility beyond this narrow range of tasks requires addressing bottlenecks such as environmental changes,” Kirstein said. “That’s why FIGURE 01 relies on superior engineering and 'Although it combines state-of-the-art deep learning, it may be overstating the progress to see it as the realization of a humanoid robot with artificial general intelligence.'
Related Posts:
