Brianna White

Administrator
Staff member
Jul 30, 2019
4,656
3,456
A team from the University of Zurich has trained an artificial intelligence system to fly a drone in a virtual environment full of obstacles before setting it loose in the real world, where it was able to weave around obstacles at 40 kph (25 mph), three times as fast as the previous best piloting software. Lead researcher Davide Scaramuzza, Director of the Robotics and Perception Group, says the work, carried out in partnership with Intel INTC -2.3%, could revolutionize robotics by enabling machines to learn virtually.
A paper describing the project, Learning high-speed flight in the wild, was published this month in the journal Science Robotics.
“Our approach is a stepping stone toward the development of autonomous systems that can navigate at high speeds through previously unseen environments with only on-board sensing and computation,” the paper concludes.
Scaramuzza’s group previously demonstrated the first autonomous drone able to beat human racing pilots around a course using sophisticated path planning with help from external sensors. The new work takes a somewhat different approach and is self-contained: the machine-learning AI effectively has an apprenticeship with an all-knowing master in the virtual world, and learns to imitate the master’s technique. (A bit like the Sparring Program in The Matrix … except that the student is an AI.)
Continue reading: https://www.forbes.com/sites/davidhambling/2021/10/25/omniscient-master-teaches-ai-drone-real-flying-skills-in-virtual-world/?sh=7ebaabb320c7
 

Attachments

  • p0005429.m05095.forbes.jpg
    p0005429.m05095.forbes.jpg
    2.8 KB · Views: 47
  • Like
Reactions: Brianna White